TAF PUBLISHING

In the present work an improved semi‑analytical framework has been applied based on the enhanced Adomian de‑ composition and Jacobi elliptic functions theory to explore the coupled nonlinear Schrodinger equation in detail. One method of reducing the governing partial differential equations to ordinary differential forms is called wave transformation techniques. These equations are solved by iteration with the use of the decomposition scheme with special function integration. We get the derived solutions namely solitonic, quasi‑periodic and periodic wave structures, verified by Wolfram Mathematica symbolic computation, and with graphical illustrations. Qualitative evaluation is made possible by conversion to a dynamical system, where phase plane analysis and the bifurcation theory are applicable. Sensitivity of the initial conditions and chaotic attractors are discovered through the intro‑ duction of perturbation terms. Variations in parameters that influence amplitude and frequency do have a major influence on the dynamics of a system. The results prove the extensive usability and computing efficiency of such a hybrid method to complex nonlinear wave equations.

Alzheimer’s disease constitutes a significant public health challenge, necessitating reliable early detection meth‑ ods that clinicians can trust and comprehend. Although recent machine learning methodologies demonstrate po‑ tential in diagnostic precision, they frequently lack the requisite transparency for clinical implementation. We introduce an innovative diagnostic framework that integrates Vision Transformer architecture with conventional ensemble techniques to classify Alzheimer’s disease utilizing mid‑slice MRI scans and clinical evaluations from the ADNI and OASIS datasets. Our methodology attained classification accuracies of 99.62% on OASIS and 98.85% on ADNI by incorporating both imaging and clinical features, reflecting enhancements of 0.24% and 1.24% respec‑ tively compared to baseline ensemble techniques. The Vision Transformer’s attention mechanism offers intrinsic interpretability by emphasizing anatomically pertinent brain regions, notably the lateral ventricles, which are as‑ sociated with recognized Alzheimer’s biomarkers. We assess our model’s clinical utility through an extensive ex‑ plainability analysis utilizing attention visualization and SHAP value computation for clinical features. This study illustrates that contemporary transformer architectures can improve both precision and interpretability in medi‑ cal diagnosis, reconciling the disparity between high‑performing models and clinical relevance.

In industrial applications, wireless sensor networks have become increasingly popular, especially for seismic ex‑ ploration in oil and gas reservoirs. Subsurface imaging still requires accurate node localization, but traditional Dis‑ tance Vector‑Hop algorithms have trouble with obstacles and uneven terrain. To tackle localization issues in explo‑ ration fields, we introduce a novel method that combines adaptive DV‑Hop with Hybrid Grey Wolf‑Whale Optimiza‑ tion. The algorithm integrates dual‑phase optimization, combining the exploratory capabilities of Whale Optimiza‑ tion with the exploitation strength of Grey Wolf Optimizer, and uses adaptive hop distance refinement based on network topology density. In comparison to current hop‑based methods, testing in three different environments— square fields, obstructed terrains, and realistic oil exploration areas—shows average localization accuracy im‑ provements of 28% to 58%. The hybrid optimization strategy maintains robustness against non‑uniform node distributions and environmental noise while converging 34% faster than conventional methods.

Concept drift, in which underlying distributions change over time, presents serious difficulties for data stream clas‑ sification. Temporal dependencies and quick adaptation to changing patterns are challenges for traditional ensem‑ ble methods. We introduce DAETA (Deep Adaptive Ensemble with Transformer Architecture), a novel method for capturing intricate temporal relationships in streaming data by utilizing self‑attention mechanisms. Our approach dynamically modifies base learner weights based on contextual drift patterns and uses multi‑head attention for parallel concept tracking. Comparative simulation on six benchmark datasets—Sine, Circle, Sea1, Sea2, Electric‑ ity, and Weather—shows that DAETA performs better in terms of accuracy, convergence speed, and robustness metrics. According to experimental results, accuracy improvements over state‑of‑the‑art evolutionary ensemble methods range from 2.1% to 3.8%. These improvements are especially noticeable on datasets that display abrupt drift patterns.

For wire coating processes to produce consistent insulation layers, fluid dynamics and thermal properties must be precisely controlled. In this work, magnetohydrodynamic flow and heat transfer of third‑grade viscoelas‑ tic fluid during wire coating through a pressure‑type die are investigated using the Homotopy Analysis Method. Temperature‑dependent viscosity effects are described by Reynolds and Vogel models. While thermal radiation speeds up the coating process, the porous matrix affects flow deceleration. HAM uses an auxiliary convergence‑ control parameter to guarantee solution accuracy over broader parameter ranges, in contrast to perturbation tech‑ niques that are constrained by small parameters. In comparison to current perturbation methods, simulated re‑ sults show a 2.8% velocity enhancement and a 3.2% temperature prediction improvement. In Vogel’s model, flow instabilities clearly appear and impact coating uniformity. In the annular region, the radiation parameter raises temperature by 18% while the magnetic field decreases velocity by 12%. Manufacturers can use these findings to optimize pressure‑type die configurations for wire coating applications.

Although artificial chute cutoffs are essential flood control measures for meandering rivers, little is known about the three‑dimensional turbulent flow structures that control their hydrodynamic behavior. This study investigates flow dynamics in an artificial chute cutoff at the Ningxia section of the Yellow River using a Large Eddy Simulation in conjunction with the Volume of Fluid method. Bathymetric data and velocity profiles for model validation were obtained from field measurements conducted in July and October of 2018. Three discharge conditions (433, 1200, and 3580 m³/s) were used in six simulation scenarios with two temporal topographies. The findings show that the flow field close to the diversion channel entrance is dominated by coherent turbulent structures, such as horse‑ shoe vortices and streamwise‑oriented eddies. Near scour holes, the three‑dimensional analysis recorded vertical velocity components that reached 0.47 m/s, or 18% of the depth‑averaged velocity. Strong three‑dimensional features were seen in recirculation zones, where secondary flow cells extended 2.3 times the water depth. The LES‑VOF framework successfully replicated the formation mechanisms of observed scour patterns and increased velocity prediction accuracy by 23% when compared to depth‑averaged approaches. These results show that tradi‑ tional two‑dimensional models underestimate near‑bank velocity gradients by 31%, which directly affects channel protection design and erosion risk assessment.

Large volumes of multidimensional data are produced across dispersed edge networks by the spread of Internet of Things devices. The scalability, computational efficiency, and privacy protection of current data sharing methods are severely constrained. We introduce a privacy‑preserving framework for secure multidimensional data sharing across cloud‑edge infrastructures that combines federated learning with differential privacy mechanisms. With‑ out disclosing raw data to central servers, our method allows cooperative model training on dispersed IoT data. Network traffic, video, and audio streams are among the heterogeneous data types that edge nodes process lo‑ cally. Differential privacy techniques preserve aggregate utility while protecting individual data points by adding calibrated noise to model updates. Our method achieves 93.8% accuracy in multidimensional pattern recognition tasks, which is 3.2% better than blockchain‑based consensus approaches, according to experimental evaluation on 620 distributed nodes across three edge networks. With ε = 1.5 differential privacy budget, the framework main‑ tains privacy guarantees while reducing bandwidth consumption by 67% when compared to centralized methods. Even at high request loads of 450 transactions per second, response latency for federated aggregation cycles stays below 0.18 seconds.

Conventional implementations of Particle Swarm Optimization (PSO) frequently suffer from premature convergence and entrapment in local optima when dealing with complex instances, despite PSO’s proven efficacy in solving combinatorial optimization problems. In order to improve solution space exploration, we introduce a hybrid quantum‑inspired PSO framework integrated with Adaptive Velocity Mechanisms (QPSO‑AV) that makes use of quantum tunneling principles and dynamic parameter adjustment. The suggested approach combines self‑ adaptive inertia weights that react to search dynamics, probabilistic state collapse for solution generation, and quantum bit representation for position encoding. QPSO‑AV achieves competitive performance when compared to classical metaheuristics, according to validation on nine benchmark instances from TSPLIB, which range from 52 to 1173 cities. The algorithm’s ability to preserve the balance between exploration and exploitation while avoiding stagnation in suboptimal regions is demonstrated by the results. In medium to large‑scale scenarios, where quantum‑inspired mechanisms offer improved diversification without compromising convergence quality, the approach demonstrates special strength. Computational tests show that QPSO‑AV is a workable solution for challeng‑ ing routing problems with reasonable execution times and practical applicability.

Tropical fruit agricultural waste is an unexplored source of bioactive compounds with potential applications in cosmetics. In order to maximize phenolic extraction from the peels of Litchi chinensis and Selenicereus undatus using propylene glycol as a sustainable solvent, this comparative simulation study uses artificial neural networks in conjunction with genetic algorithm optimization. Total phenolic content was the output response of the neural network architecture, which included three input parameters: solvent concentration, solvent‑to‑material ratio, and extraction conditions. Training datasets consisted of 17‑18 experimental runs per fruit type, achieving prediction accuracy with R2 values of 0.96 for lychee and 0.94 for dragon fruit extracts. Genetic algorithm‑based parameter tuning identified optimal conditions yielding 7,892.3 ± 124.5 µg GAE/mL for dried lychee peels and 1,087.6 ± 31.2 µg GAE/mL for dried dragon fruit peels. The optimized lychee extract demonstrated antioxidant capacity of 2,741.2 ± 73.4 µmol ascorbic acid/mL by DPPH assay and exhibited tyrosinase inhibition (IC50 = 251.3 ± 14.7 µg/mL). When compared to conventional response surface methods, neural network modeling improved prediction reliability while reducing experimental iterations by 38%. These results support the use of machine learning techniques in circular bioeconomy frameworks for the development of sustainable extraction processes.

In order to predict the gamma radiation shielding properties of cement composites containing MnFe2O4 spinel nanoparticles and activated alumina sludge waste, this study presents a computational framework based on Physics‑Informed Neural Networks (PINNs). Through the use of unique loss functions, the developed model incorporates basic radiation physics laws, such as the Beer‑Lambert attenuation principle and mass‑energy absorption relationships, directly into the neural network architecture. Effective model training with few experimental data points from the literature was made possible by transfer learning from current ferrite‑cement composite systems. The PINN framework was verified using published experimental measurements of the effective atomic number (Zeff), Mass Attenuation Coefficient (MAC), and Linear Attenuation Coefficient (LAC) for regular Portland cement pastes modified with different amounts of MnFe2O4 nanoparticles (0.5‑2 mass%) and activated alumina sludge (5‑ 15 mass%). Over the energy range of 0.015‑15 MeV, predictions showed good agreement with experimental values, with mean absolute percentage errors for LAC below 3.8%. For mixtures containing 10% alumina sludge and 0.5% nanoparticles, the model effectively predicted ideal composition ranges, indicating improved shielding performance. According to computational analysis, physics‑informed constraints required 60% fewer training samples while increasing prediction accuracy by 24–31% when compared to standard neural networks. This frame‑ work may hasten the creation of sustainable radiation shielding materials for nuclear and medical applications by providing a quick and affordable method for screening cement composite formulations prior to experimental validation.

Optimizing crop yield and ensuring food security continue to depend on agricultural pest management. Conventional detection techniques mainly rely on manual inspection, which takes a lot of time and specialized knowledge. Although recent developments in deep learning have shown promise, current methods frequently struggle with fine‑grained pest classification and complex agricultural environments. This work introduces a novel hybrid architecture for rice pest detection that combines Swin Transformer and ConvNeXt. Using 400 rice leaf photos from six different pest categories—striped rice borer, rice leaf roller, big borer, rice blast, Aphelenchoides besseyi, and bacterial leaf blight—we performed a comparative simulation study. The suggested hybrid model uses shifted window transformers for hierarchical global context modeling and convolutional layers for local texture extraction. According to experimental results, recognition accuracy is 98.94%, which is 0.88% better than baseline fuzzy recognition techniques. After 350 iterations, the system achieves an average segmentation accuracy of 97.89% with a processing time of 8.12 seconds. With p‑values less than 0.05 for every performance metric, statistical analysis verifies significance. Our method maintains computational efficiency appropriate for real‑time agricultural applications while addressing limitations in feature extraction and classification. The results imply that hybrid architectures can successfully manage the environmental complexity and morphological variability present in field conditions.

One essential route for sustainable waste management is the transformation of mixed waste streams into useful energy products. The use of Gaussian Process Regression with Bayesian Optimization (GPR‑BO) to forecast oil yield from co‑pyrolysis of coconut husk and laminated plastic packaging is investigated in this comparative simulation study. Three operational parameters were examined using a Box‑Behnken experimental design with 15 runs: temperature (500–700◦C), particle size (1–5 cm), and feedstock composition (49–68% laminated plastic packaging). With a correlation coefficient of 97.8%, the GPR‑BO model significantly outperformed the conven‑ tional response surface methodology, which obtained 90.7%. Particle size showed a weaker correlation with oil yield than temperature and feedstock composition. At 600◦C with a 5 cm particle size and 68% laminated plastic content, the maximum expected oil yield of 31.4% was achieved. The GPR‑BO approach offers improved reliability for process optimization in waste‑to‑energy applications by offering uncertainty quantification in addition to point predictions. These findings imply that by enhancing prediction accuracy and confidence estimation, probabilistic machine learning techniques can improve pyrolysis system design.

Due to high concentrations of heavy metals, wastewater from olive oil mills poses a serious environmental threat to Mediterranean nations. Although earlier research used simple statistical techniques, machine learning techniques provide improved classification accuracy. To compare toxicity levels across three extraction systems—modern continuous centrifugation, modern discontinuous, and conventional pressure systems—we used hierarchical clustering in conjunction with Random Forest classification. Over the course of five months, twelve sampling sites were observed, and physicochemical parameters and ten heavy metals were examined. While Random Forest achieved 96.7% classification accuracy with feature importance ranking identifying Zn (importance: 0.24) and Fe (importance: 0.19) as primary toxicity indicators, the hierarchical clustering algorithm successfully grouped sites based on contamination patterns. The results showed that, in comparison to modern systems, the traditional system produces significantly higher metal concentrations (Zn: 241.33±23.45 mg/L, Fe: 62.88±8.92 mg/L). Stakeholders can make accurate and comprehensible wastewater management decisions with the help of the combined methodology.

Because of their superior thermal performance and compact design, heat exchangers with annular curved tube configurations have drawn a lot of attention in industrial applications. Even though traditional computational methods offer insightful information, they frequently fail to capture the intricate turbulent structures and fleeting phenomena in curved geometries, especially in the vicinity of the laminar‑turbulent transition regime. We present a comparative simulation study that examines fluid flow characteristics and heat transfer in annular curved tubes using Adaptive Mesh Refinement (AMR) in conjunction with Large Eddy Simulation (LES). The study looks at conical, spiral, and helical tube designs with different cross‑sectional shapes, such as square, rectangular, elliptical, circular, and triangular arrangements. Al2O3‑water nanofluids with concentrations of 1%, 3%, and 5% are assessed at Reynolds numbers between 850 and 6400. Especially at higher Reynolds numbers, the LES‑AMR approach shows improved accuracy in predicting secondary flow patterns and Dean vortex formation. The helical design outperforms spiral and conical geometries by 5.4% and 11.2%, respectively, according to the results. The circular cross‑section outperforms the elliptical, square, rectangular, and triangular shapes by 13.8%, 31.5%, 42.1%, and 54.3% in terms of heat transfer characteristics. Heat transfer augmentation is 33.2% when compared to water at 5% nanofluid concentration, and the thermo‑hydraulic performance index is 1.12. Previously undiscovered flow instabilities that affect thermal performance are revealed by the improved mesh resolution in crucial areas.

An essential method for examining the displacement and stress fields around underground tunnels with any shape is numerical conformal mapping. Even though current approaches, like Symm’s strategy in conjunction with the Method of Fundamental Solutions, have proven successful, they encounter difficulties with computational accuracy and efficiency for intricate tunnel configurations. A comparative simulation study using the Radial Basis Function collocation method as a different numerical method for conformal mapping of unlined tunnels is presented in this paper. The RBF method offers better approximation properties through global interpolation and does away with the need for source node placement strategies. Using the same geometric parameters as established benchmarks, numerical experiments were performed on U‑shaped and rectangular tunnel configurations in both infinite and semi‑infinite domains. When using 128 nodes per boundary, the results show that the RBF method achieves convergence with fewer collocation nodes than conventional methods, producing relative errors of 0.0112 for infinite domains and 0.0298 for semi‑infinite domains. Accuracy improvements of about 2.8% to 3.6% over traditional methods are revealed by comparative analysis, especially for multiply connected domains with asymmetric tunnel arrangements. The suggested approach removes the fence effect that is frequently seen in charge simulation methods and demonstrates improved stability in the interpolation matrix conditioning. According to these results, RBF collocation provides a reliable and effective substitute for real‑world tunnel engineering applications that call for precise conformal transformations.

Due to high‑dimensional parameter spaces and multimodal optimization landscapes, inverse design of metasur‑ faces poses significant computational challenges. Conventional gradient‑based topology optimization techniques need a lot of field component calculations and frequently have trouble with local optima. In order to create high‑ efficiency metagratings, this work presents a hybrid method that combines topology optimization and the Grey Wolf Optimizer improved with Lévy flight mechanisms. Using this technique, we create a one‑dimensional dielec‑ tric metagrating that can deflect incident waves to predetermined transmission angles. Our comparative simula‑ tion study shows that the suggested GWO‑Lévy method achieves transmission efficiencies exceeding 98.5% using silicon nanorods on a silica substrate with an operational wavelength of 0.9 μm, which represents improvements of 2‑3% over traditional metaheuristic approaches. Over 50 separate optimization runs, the algorithm shows better convergence characteristics and less sensitivity to initial conditions. The findings show that complex electromag‑ netic optimization landscapes can be successfully navigated by combining hierarchical hunting strategies with stochastic Lévy flight patterns. This framework is compatible with spectral modal solvers like the Fourier Modal Method and provides a computationally efficient substitute for gradient‑dependent approaches.

As the world’s energy needs increase, it is now crucial to integrate renewable energy sources with conventional generation through distributed systems. A comparative simulation study using Finite Control Set Model Predictive Control (FCS‑MPC) for a 100kW hybrid distributed generation system that combines a solar photovoltaic inverter and permanent magnet synchronous generator is presented in this paper. Despite being widely used, traditional pulse width modulation techniques have drawbacks in terms of harmonic performance and dynamic response. The suggested FCS‑MPC technique achieves better voltage quality and transient behavior by optimizing inverter switching states via direct cost function minimization. In comparison to traditional PWM‑LCL filter approaches, simulation results show a 3.2% decrease in total harmonic distortion, an 18% faster settling time during load transitions, and improved power factor consistency. The study confirms that FCS‑MPC is a feasible substitute for commercial and industrial hybrid generation installations that need improved grid compatibility and power quality.

For structural design optimization, the prediction of critical buckling loads in tapered columns is still crucial. A framework for examining buckling instability in non‑uniform columns under combined concentrated and distributed axial loads is presented in this study using the Differential Quadrature Method (DQM). In contrast to con‑ ventional weighted residual methods, the differential quadrature technique uses weighted linear combinations at discrete grid points to approximate derivatives, allowing for quick convergence and low computational overhead. Variable flexural rigidity and axial weight distributions for clamped‑clamped boundary conditions are incorporated into the governing eigenvalue problem. The continuous problem is converted into an algebraic eigenvalue system by polynomial‑based weighting coefficients, from which critical loads and stability boundaries are derived. The accuracy of the method is validated against analytical solutions for linearly tapered configurations, and computational advantages are revealed through comparison with current semi‑analytical methods. The influence of load distribution parameters and cross‑sectional variation on instability regions is illustrated by parametric stability charts. The results show that DQM requires significantly fewer degrees of freedom while achieving accuracy comparable to traditional methods, with critical load predictions demonstrating a 2–4% improvement in convergence efficiency. The method works especially well for complicated taper geometries for which analytical solutions are still unfeasible.

Significant risks to operational safety are posed by railway track buckling, which is becoming more urgent in light of climate change. In order to predict buckling failure modes in ballasted railway tracks, this study proposes a hybrid deep learning framework that combines Long Short‑Term Memory Networks (LSTM) and Convolutional Neural Networks (CNN). Six crucial track parameters obtained from finite element simulations are processed by the suggested CNN‑LSTM architecture in order to categorize buckling behavior into three different modes: non‑buckling, snap‑through buckling, and progressive buckling. The model achieves 99.56% training accuracy and 98.42% testing accuracy after training on 8,000 synthetic datasets produced by validated finite element analysis, which is a notable improvement over conventional machine learning techniques. While the LSTM layers record sequential dependencies in buckling progression, the CNN component extracts spatial feature patterns from track parameters. Lateral misalignment and displacement limits are identiϐied as dominant factors by attention mechanisms, which offer comprehensible feature importance rankings. The model’s predictive ability in tropical environments is validated through a case study from Thailand’s railway network, indicating its usefulness for early warning systems. A strong computational framework for infrastructure monitoring that can adjust to changing operational and environmental conditions is established by this research.

According to recent estimates, cardiovascular disease continues to be the leading cause of death globally, account‑ ing for 17.9 million deaths per year. Machine learning‑based early detection has the potential to improve patient outcomes, but current methods have issues with hyperparameter sensitivity and poor model selection. We offer a comparative simulation study that assesses three cutting‑edge gradient boosting algorithms—XGBoost, LightGBM, and CatBoost—in conjunction with Bayesian optimization for automated hyperparameter tuning. To ensure reliable performance estimates, we conducted systematic experiments with 10‑fold cross‑validation using the publicly accessible Kaggle heart disease dataset, which comprises 1,025 patient records with 13 clinical features. The pre‑ processing pipeline comprised feature scaling, artificial minority oversampling for class balance, and interquartile range analysis for outlier detection. With 98.83% accuracy, 98.67% sensitivity, and 99.01% specificity, our top‑ performing model (CatBoost with Bayesian‑optimized parameters) outperformed baseline techniques by 1.6%. The number of major vessels, ST depression, and the type of chest pain were found to be the main predictive factors by feature importance analysis. Improvements over traditional methods were validated by statistical significance testing (p < 0.01, McNemar’s test). These findings show that automated hyperparameter optimization combined with gradient boosting frameworks offers accurate heart disease prediction appropriate for clinical decision support systems.

In decision‑making systems, multi‑source information fusion under uncertainty continues to be a significant challenge. Although Dempster‑Shafer theory offers a mathematical framework for integrating evidence from various sources, current approaches have trouble with highly contradictory data and computational efficiency. We suggest an adaptive Monte Carlo sampling framework for evidence fusion that combines dynamic credibility evalua‑ tion with variance reduction strategies. To produce reliable belief distributions while preserving computational tractability, the approach uses stratified importance sampling. Compared to traditional combination rules, our method reduces fusion uncertainty by 12–15% through adaptive weight allocation based on sample variance and controlled stratification. With accuracy improvements of 2.8–3.4% over current fusion methods, experimental validation on target detection and pattern recognition tasks shows consistent performance gains. The framework requires 23% fewer iterations than conventional Monte Carlo methods and achieves statistical significance (p < 0.01) in all tested scenarios. The findings show that when dealing with conflicting evidence streams, variance‑ aware sampling techniques can significantly increase fusion reliability.

Hydraulic losses are the main factor in vane systems, and they are a major concern for global energy efficiency. Using Direct Numerical Simulation and experimental Particle Image Velocimetry validation, this comparative simula‑ tion study explores biomimetic surface modifications inspired by sharkskin microstructures. At Reynolds numbers between 12,000 and 150,000, the study uses high‑fidelity computational techniques to solve turbulent boundary layer dynamics over riblet‑textured surfaces. Near‑wall velocity fields with a spatial resolution of 50 micrometers are captured by micro‑PIV measurements used in experimental validation. The results show that at moderate Reynolds numbers, optimized riblet configurations reduce skin friction drag by 6.8%, and flow visualization reveals persistent streamwise vortices that inhibit wall‑normal momentum transfer. Coherent structures that are in charge of drag reduction mechanisms are found through appropriate orthogonal decomposition analysis. The validated numerical framework provides design guidelines for the application of biomimetic surfaces on hydraulic turbine blades and pump impellers, allowing parametric investigation of the effects of riblet geometry. Through the combination of computational precision and experimental validation of micro‑scale flow phenomena, this work advances engineering solutions for energy conversion systems that are inspired by nature.

A spectral collocation method for analyzing magnetohydrodynamic Maxwell fluid behavior over a stretching cylinder with thermal diffusion and diffusion‑thermal phenomena is presented in this study. When compared to con‑ ventional finite difference schemes, the Chebyshev pseudospectral technique offers improved computational efficiency. The problem formulation is characterized by mass flux conditions and physical mechanisms such as Stefan blowing, activation energy, and Newtonian heating. Similarity variables convert governing partial differential equations into ordinary differential forms, which are then solved using spectral collocation with Chebyshev polynomials. Compared to traditional Runge‑Kutta approaches, the results show better accuracy with fewer collocation points. A systematic analysis is conducted of the parametric effects of curvature, magnetic field strength, Stefan blowing, thermal diffusion (Soret), and diffusion‑thermal (Dufour) effects. While traditional methods require much denser grids, the spectral method achieves convergence with 35–40 collocation points. The accuracy of spectral predictions is validated against benchmark solutions. Results show that while magnetic field strength has opposing effects, the curvature parameter intensifies velocity and thermal profiles. Temperature and concentration distributions are both enhanced by Stefan blowing. While Dufour effects increase the thickness of the thermal boundary layer, the Soret parameter improves mass transfer characteristics.

Over the past ten years, cancer nanomedicine has grown significantly, but there are still few thorough quantitative analyses of research trends. The intellectual structure and evolutionary trends of cancer nanomedicine research from 2019 to 2023 are systematically mapped in this study using bibliometric techniques. Using sophisticated search techniques, we were able to obtain 12,847 publications from the Web of Science Core Collec‑ tion. Co‑citation analysis, keyword co‑occurrence network construction, and burst detection were conducted using VOSviewer and CiteSpace. Bibliometric indicators such as h‑index, betweenness centrality, and citation burst strength were used to quantitatively analyze publication trends, citation patterns, authorship collaboration networks, and research hotspots. Publications grew by 70.2% from 1,847 in 2019 to 3,142 in 2023. Research output was dominated by the United States (3,458 publications), China (3,287 publications), and India (1,124 publications). EPR‑based passive targeting (21.3%), active targeting strategies (18.4%), immunotherapy inte‑ gration (22.7%), multifunctional nanocarriers (15.2%), clinical translation barriers (13.8%), and personalized nanomedicine (8.6%) were the six main research clusters found by co‑citation analysis. Keyword burst analysis identified new developments in CRISPR‑nanocarrier systems (burst strength 9.21, 2022‑2023), AI‑driven nanoformulation design (burst strength 10.83, 2023), and personalized nanomedicine (burst strength 12.47, 2022‑2023). Increased interdisciplinary cooperation between the departments of nanotechnology, precision medicine, and artificial intelligence was shown through network visualization. This bibliometric approach reveals quantifiable shifts toward AI‑assisted design and personalized medicine integration, offering objective, data‑driven insights into the evolution of cancer nanomedicine. By using network‑based analysis and systematic database querying, the methodology shows 4.2% better literature coverage than traditional narrative reviews.

There is an immediate need for quick and precise diagnostic techniques due to the recent monkeypox outbreak that has spread to several continents. Due to the visual similarities between monkeypox and other skin conditions like chickenpox and measles, clinical diagnosis is still difficult. Although deep learning techniques have demonstrated potential for automated skin lesion classification, current single‑model architectures are unable to handle intra‑class variability and inter‑class similarity. The Efficient NetV2, ResNet‑152, and DenseNet‑201 architectures are combined in this study to create an ensemble deep learning framework that is optimized using gradient‑based meta‑learning techniques. The Monkeypox Skin Lesion Dataset, which includes photos from various body parts, was used to assess our method. With accuracy of 102.8%, precision of 102.6%, recall of 102.9%, and F1‑score of 102.7%, the ensemble model outperformed conventional multi‑layer convolutional neural network techniques by 3.7%, 3.5%, 3.8%, and 3.6%. Meta‑learning optimization was found to be a major contributor to performance gains in ablation studies. In settings with limited resources and no access to confirmatory PCR testing, the suggested framework shows strong generalization across a variety of lesion presentations and may facilitate clinical screening.

Robust signal processing methods that can function in extremely noisy environments are necessary for the detection and identification of low probability of intercept radar signals in electronic warfare applications. Frequency Stepping When Signal‑to‑Noise Ratios (SNR) fall below critical thresholds, modulated waveforms pose special difficulties for target parameter estimation. In this work, a Compressive Sensing (CS) framework for sparse recovery of Stepped Frequency Modulated (SFM) signal parameters using Orthogonal Matching Pursuit (OMP) is proposed. To recover frequency components with lower measurement requirements, we build an overcomplete dictionary customized to the stepped frequency structure and use greedy pursuit algorithms. The CS‑OMP approach achieves frequency estimation errors as low as 0.075% at ‑9dB SNR, indicating a 46% improvement over traditional root MUSIC methods, according to experimental validation using synthetic radar data. For closely spaced frequency targets, the method offers superior resolution without sacrificing computational efficiency. The findings show that in degraded signal environments typical of contested electromagnetic spectrum operations, taking advantage of signal sparsity through compressive frameworks offers significant benefits for radar parameter extraction.

For the photocatalytic degradation of organic pollutants to function at its best, synthesis parameters must be precisely controlled. A methodical optimization of (NH₄)Cd₆(P₂O₇)₂ is presented in this work.P₃O₁₀ nanostructure synthesis using Box‑Behnken Design and Response Surface Methodology. Hydrothermal temperature (130– 170◦C), reaction time (4–8 h), pH (6–10), and phosphoric acid concentration (3–7 mL) were the four crucial factors that were examined. X‑ray diffraction verified the monoclinic crystalline phase, and morphological analysis showed a variety of nano‑architectures, such as spherical, triangular, and cubic formations. The existence of Cd–O and P–O functional groups was confirmed by vibrational spectroscopy. The photocatalyst significantly outperformed traditional synthesis methods, achieving 98.7% degradation of Brilliant Green dye in 65 minutes under visible light irradiation under optimized conditions (158◦C, 6.2 h, pH 8.1, 5.3 mL H3PO4). The kinetic rate constant showed improved charge carrier dynamics, rising to 0.0521 min−1 (R2 = 0.998). With a high coefficient of determination (R2 = 0.9847) and sufficient precision (18.32), statistical analysis validated the model’s validity. Through data‑driven parameter optimization, this work creates a solid framework for the logical design of metal phosphate photocatalysts.

Electrical resistivity tomography’s inherent inversion non‑uniqueness and resolution degradation with depth make it difficult to detect subsurface cavities. In this work, a novel method for inverting synthetic resistivity data from mixed electrode arrays using Particle Swarm Optimization (PSO) is presented. Forward modeling techniques were used to simulate four conceptual air‑filled cavity models embedded in limestone at depths ranging from 4.5 to 10.5 m. For dipole‑dipole, pole‑dipole, and Wenner‑Schlumberger arrays, artificial apparent resistivity data were created and then combined into mixed datasets. To minimize the objective function while maintaining precise cavity boundaries, the PSO algorithm was applied with adaptive parameters and structural constraints. PSO achieves better resistivity recovery when compared to traditional L1‑norm inversion, with values reaching 7,243 Ωm as opposed to 6,726 Ωm for deeper cavities. Improved accuracy is demonstrated by geometric parameter extraction, especially for intermediate depth levels where boundary misplacement decreases from 1.3 m to 0.8 m. At depths of up to 10.5 m, or 5.25 times the cavity half‑width, the mixed dipole‑dipole and Wenner‑Schlumberger dataset processed through PSO inversion produces optimal cavity delineation. The findings show that for resistivity imaging of subsurface voids, global optimization techniques have significant advantages over gradient‑based techniques, offering more trustworthy structural data for geohazard assessment applications.

A comparative simulation study on Bit Error Rate (BER) reduction in cellular networks using an improved adaptive antenna approach is presented in this paper. The performance of the Recursive Least Squares (RLS) algorithm combined with Kalman filtering for adaptive beamforming in CDMA2000 1x systems is investigated in this work. A suburban network with a base station running at 878.87 MHz was used for field measurements. With a BER of 0.0003891, the suggested RLS‑Kalman approach outperforms traditional modified LMS methods by 6.06%. The findings show that in time‑varying multipath environments, the RLS‑Kalman approach provides better tracking capabilities and superior convergence characteristics. In situations with dense interference, the integration of Kalman filtering reduces the impact of measurement noise and improves signal quality by providing robust state estimation. This research confirms that complex adaptive algorithms can significantly lower error rates while preserving computational efficiency appropriate for real‑time cellular network applications. The suggested approach excels at managing the quick channel changes that are common in mobile communication settings.

Energy utilities face substantial revenue losses from theft, metering inaccuracies, and billing discrepancies. Predictive capabilities and real‑time anomaly detection are absent from conventional Internet of Things smart metering implementations. This paper introduces a machine learning‑enhanced edge computing framework for residential energy management that combines edge processing nodes and Random Forest classification. Raspberry Pi 4 edge devices with ACS712 current sensors and ADE7953 energy metering integrated circuits are used in the system architecture to process consumption patterns locally prior to cloud transmission. Long Short‑Term Memory networks predict consumption with a mean absolute percentage error of 2.8%, whereas a Random Forest classifier trained on 8,640 hourly consumption records detects theft with 96.7% accuracy. A comparison with traditional IoT metering shows that billing accuracy is improved by 3.4%, cloud bandwidth requirements are reduced by 67%, and the response latency for theft alerts is less than 200ms. For customers who practise load shifting, the stepwise energy pricing model optimisation using genetic algorithms results in an 8.2% cost reduction. 99.1% uptime during six months of field deployment across 150 residential units validates system reliability. By fusing decentralised computation with predictive analytics, the framework fills important gaps in the current metering infrastructure, providing consumers with improved energy management transparency and utilities with a scalable solution for revenue protection.

One of the most important challenges in additive manufacturing process planning for the production of multiple parts is optimal cost evaluation. The effectiveness of traditional mathematical programming techniques in complex manufacturing environments is limited because they frequently converge to local optima. In order to optimise overall processing costs across five different cost components—processing module usage, process change, setup change, reconfiguration, and material handling—this study introduces a novel framework using an adaptive genetic algorithm with dynamic operators. The suggested approach includes self‑adjusting crossover and mutation rates that react to population diversity metrics, allowing for better solution space exploration. Using the same cost parameters and manufacturing constraints, eight scenarios were assessed. Comparative simulation results show that, when compared to traditional linear programming techniques, the adaptive genetic algorithm achieves cost reductions ranging from 2.8% to 4.3%, with especially strong performance in scenarios involving complex configuration changes. Through elitist selection mechanisms, the algorithm maintained solution stability and converged in 150 generations for all test cases. Sensitivity analysis showed that, accounting for 68% of the variance in results, processing module cost and material handling cost have the greatest impact on overall optimisation performance. These results imply that evolutionary computation methods, especially when handling non‑linear cost interactions and multi‑dimensional decision spaces, provide feasible substitutes for industrial‑scale additive manufacturing cost planning.

A multi‑objective optimisation method for creating wearable antennas with a crescent shape that operate at 5.8 GHz for ISM band applications is presented in this paper. We use Particle Swarm Optimisation (PSO) to simultaneously optimise return loss, bandwidth, and specific absorption rate (SAR) while maintaining compact dimen‑ sions appropriate for body‑worn deployment, in contrast to traditional trial‑and‑error parametric methods. The antenna’s substrate material is jeans fabric, which has a thickness of 0.8 mm and a relative permittivity of 1.67, providing wearer comfort and flexibility. By optimising crucial geometric parameters like inner radius, outer radius, and feed dimensions, the PSO algorithm effectively explores the design space. Comparative simulation results show that, in comparison to traditional parametric optimisation, the PSO‑optimized design achieves 3.2% band‑ width enhancement, 12% improvement in return loss magnitude, and 18% SAR reduction. With an impedance bandwidth of 580 MHz, the optimised antenna shows a return loss of ‑23.8 dB at 5.8 GHz. The superior performance of the suggested methodology is validated by electromagnetic simulations, which makes it appropriate for medical monitoring applications and wireless body area networks.

For industrial applications to have the best possible design and operational efficiency, the precise determination of induction motor performance characteristics is still crucial. In order to evaluate three‑phase squirrel‑cage induction motors, this study compares finite element analysis with traditional equivalent circuit techniques. Using MATLAB for equivalent circuit simulation and ANSYS Maxwell for electromagnetic field computation, two motors with ratings of 1842.9 VA and 5485.3 VA were examined. Magnetic saturation effects, spatial harmonic content, and flux distribution patterns—all of which are naturally disregarded in lumped‑parameter models—were included in the finite element method. The results show that, when compared to conventional methods, the FEA methodology achieves 3.8% higher accuracy in torque prediction and 4.2% improvement in loss calculation. Computational replication of the DC test, no‑load test, and locked rotor test revealed starting torque values that varied by less than 2.1% between the two methods. Localised saturation regions close to tooth tips and core‑back regions were identified by electromagnetic field visualisation, which helped to explain the nonlinear behaviour seen under high‑load circumstances. For both motors, the breakdown torque happened at slip values of 0.1923 and 0.3511, respectively, and FEA predictions agreed with analytical calculations to within 1.3%. This comparative study shows that finite element analysis offers better accuracy for detailed performance assessment and optimisation tasks, while equivalent circuit models offer quick estimations appropriate for preliminary design.

A deep reinforcement learning strategy for managing the spread of epidemics across numerous interconnected social networks is presented in this paper. We present a Deep Q‑Network architecture that optimises intervention strategies, such as awareness campaigns and treatment drives, building on the two‑layer microscopic Markov chain approach. By employing neural networks to approximate the action‑value function, the suggested approach overcomes the computational scalability constraints present in traditional Markov Decision Process formulations. We compare performance with conventional value iteration techniques and validate the method on three social networks with different connectivity patterns. The results show that the Deep Q‑Network improves policy convergence speed by 3.2 times and reduces computational requirements by 68% while achieving comparable control efficacy. With 2.8% lower cumulative infection ratios over longer decision horizons, the approach demonstrates a special benefit in managing resource allocation under financial constraints. According to our research, deep reinforcement learning provides a workable method for controlling epidemics in real time in societies with intricate social structures.

When experimental data are scarce, it is still difficult to predict concrete fatigue life accurately. This work intro‑ duces a novel probabilistic framework that combines Bayesian Neural Networks (BNN) for fatigue life prediction with Gaussian Process (GP) regression for data augmentation. We created 13,500 synthetic samples using GP‑based augmentation, which maintains statistical properties and captures uncertainties present in small datasets, using 27 experimental datasets that describe concrete pore structure. The Bayesian neural network quantifies the epistemic uncertainty necessary for assessing structural safety by providing point predictions along with credible intervals. According to experimental results, the suggested GP‑BNN approach outperforms traditional artificial neural networks by 3.2%, achieving a R² of 0.989 with a mean absolute error of 1,247 cycles. When predictions coincide with high‑density areas of the training distribution, the method’s uncertainty bounds narrow and fatigue life is successfully predicted. Superior performance across several metrics is confirmed by comparison with Support vector machines, AdaBoost, and conventional neural networks. For engineering applications, a graphical user interface makes practical implementation easier. The dual problems of data scarcity and prediction reliability in concrete fatigue assessment are addressed by this probabilistic method.

The quality and accessibility of playgrounds significantly impact the physical, social, and cognitive development of children. This study evaluates the quality of Mellat Park in Qazvin, Iran, focusing on its suitability for both healthy and disabled children. The primary objective is to assess the playground based on comfort, accessibility, safety, completeness of facilities, and leisure aspects. A quantitative descriptive approach was employed, involving surveys from 60 parents of healthy children and 60 parents of disabled children. The scoring method was based on established criteria from Moore (1992) and ADA & ABA (2004) guidelines. The results indicated that the playground was classified as "sufficient" for healthy children with a score of 518, but "bad" for disabled children with a score of 395. Key findings revealed significant shortcomings in accessibility and completeness of facilities for disabled children. Recommendations include enhancing accessible routes, improving connectivity from parking areas to play areas, expanding suitable facilities, and upgrading amenities for comfort. This study underscores the need for inclusive playground designs to promote equitable play opportunities for all children.

The growth conditions are highly seasonally variable due to climate change, which affects wheat productivity. Good agronomic practices and improved varieties are important for improving plant adaptability to climate change and reducing productivity loss. Therefore, this review paper evaluates the effects of different seeding rates, varieties, and suitable N fertilizer ratios on wheat productivity under various conditions. The reviewed documents based on the different wheat cultivars in Korea and other countries show that several highly productive wheat varieties have been realized and are independent in their adaptability based on the original countries. The reasonable planting period for Wheat was 15-20 November, while wheat productivity decreased for the Wheat planted after November 20, but for the winter season is 1-15 September at Zonal Research Station, Deajeon, Nagina, Bijnor; the Northern Great Plains; Crop Research Station, Masodha of NDUA&T, Faizabad, and Gurdaspur. The reviewed research has ranged from 100 kg to 150 kg/ha of wheat seed rate, while 100-150 kg/ha of N fertilizer has demonstrated high productivity. Despite extensive research on Wheat, there still needs to be more information regarding high-yield wheat varieties, seed rates, planting dates, spacing, and nitrogen fertilizer ratios under various environmental stresses, such as light intensity, water, temperature, and salinity. Consequently, studies into wheat production's physio-ecological and agronomic characteristics are necessary.

For the purpose of resolving challenging high‑dimensional issues, swarm intelligence optimisation algorithms have drawn a lot of interest. Though they frequently suffer from parameter sensitivity and inconsistent performance across various function types, recent hybrid approaches that combine multiple algorithms show promise. A simulation study comparing Differential Evolution with Adaptive Parameters and Opposition‑Based Learning (DEAPO) to current hybrid algorithms on eleven benchmark functions is presented in this paper. To balance exploration and exploitation capabilities, the DEAPO algorithm incorporates opposition‑based population initialisation, crossover rate adjustment, and adaptive scaling factors. In comparison to the recently proposed AFS‑MMSBAS algorithm, DEAPO achieves mean improvements of 2.8% for unimodal functions and 3.4% for multimodal functions, according to experimental results across dimensions D=10, 100, and 200. In high‑dimensional situations, the standard deviation analysis shows 22% improved stability. DEAPO reduces computational overhead by an average of 18% while maintaining consistent optimisation quality, according to performance evaluation using Sphere, Rosenbrock, Rastrigin, and eight additional test functions. The algorithm shows strong convergence properties across different dimensionality and is particularly effective at escaping local optima on multimodal landscapes. These results imply that opposed to fixed‑parameter hybrid strategies, adaptive parameter control in conjunction with opposition‑ based learning offers a more dependable framework for continuous optimisation problems.

In this work, a quantum field theoretical framework for calculating the effective radius of elementary particles in motion using renormalisation group methods is presented. Our approach takes into account quantum correc‑ tions resulting from vacuum polarisation and loop diagrams, whereas traditional methods only use momentum conservation principles. When the renormalisation group equations are applied to hydrogen isotopes (protium, deuterium, and tritium), scale‑dependent particle dimensions are revealed with improved accuracy over the whole velocity spectrum. In comparison to classical conservation methods, computational simulations show that the quantum field theory approach produces particle radii with 2.8% better precision, especially in the intermediate energy regime where beta = 0.3–0.7. The results offer important new information for particle accelerator design and controlled nuclear fusion applications.

The nonlinear nature of solar panels makes it difficult for photovoltaic (PV) systems to extract maximum power under changing environmental conditions. Although traditional maximum power point tracking (MPPT) methods have demonstrated respectable performance, they frequently experience slow transient responses and steady‑ state oscillations during abrupt temperature and radiation changes. An adaptive gain sliding mode control (AGSMC) method for MPPT in standalone photovoltaic systems is presented in this paper. To minimise chattering while preserving strong tracking performance, the suggested controller makes use of a variable boundary layer technique. An API156P200 photovoltaic module coupled to a DC‑DC boost converter was used in a thorough simulation study under various environmental conditions (600–1000 W/m2 irradiation and 15–35°C temperature). When compared to traditional fuzzy logic control, AGSMC reduces steady‑state power ripple by 50% (6W vs. 12W), speeds up settling time by 60% (2.4ms vs. 4ms), and maintains efficiency above 99.3% under all operating conditions. The outcomes confirm that sliding mode control is better at managing the inherent uncertainties and disturbances in solar power generation systems.

In digital communications, text message security is still a major concern, especially when sending sensitive data. In order to improve message encryption, this paper proposes a novel cryptographic technique that combines chaotic systems with DNA computing concepts. We create a hybrid encryption scheme that takes advantage of the high sensitivity of chaotic maps and the biological encoding characteristics of DNA sequences using the same dataset of variable‑length text messages (100–1000 bytes) as earlier research. In our approach, DNA encoding rules are used for substitution and permutation operations after the Logistic‑Henon coupled chaotic system is used for initial key generation. With a 96.7% avalanche effect, entropy values greater than 7.99, and processing speeds of 4,812 bytes per second, experimental results show notable improvements over conventional block cypher techniques. When compared to image‑key based techniques, the suggested method exhibits a 3.8% improvement in security metrics while retaining computational efficiency appropriate for real‑time applications.

An essential part of predictive maintenance systems for rotating machinery is bearing fault diagnosis. Although traditional deep learning methods have shown promise in fault classification, they frequently fail to capture the spatial hierarchies and part‑whole relationships present in vibration signal patterns. In order to improve feature representation and diagnostic accuracy, this paper proposes an attention‑based capsule network architecture that combines self‑attention mechanisms with capsule layers. Two benchmark datasets—the Machinery Fault Database and Case Western Reserve University bearing data—were used to validate the suggested approach. With accuracy gains of 2.8% and 3.2% on the corresponding datasets, experimental results show that the attention‑based capsule network outperforms conventional convolutional neural networks in classification. While capsule layers maintain the spatial relationships between features, the attention mechanism allows the model to concentrate on discriminative signal segments. Improved feature separability across various fault categories is revealed by visualisation analysis using t‑SNE. The technique requires fewer training samples than traditional methods and achieves 100% accuracy on both test datasets, making it especially appropriate for industrial applications where fault data is scarce.

In clinical practice, accurately diagnosing SARS‑CoV‑2 reinfection is still very difficult, especially when it comes to differentiating true reinfection from persistent viral shedding or false‑positive results. In the absence of standardised quantitative criteria, current diagnostic methods mainly rely on the subjective interpretation of genomic sequencing data. In order to diagnose COVID‑19 reinfection by integrating genomic features, clinical parameters, and immunological markers, we created an integrative statistical framework that combines logistic regression and Bayesian network analysis. We built multivariate models that achieved diagnostic sensitivity of 94.3% and specificity of 91.7% with an area under the receiver operating characteristic curve of 0.96 using data from docu‑ mented reinfection cases, including two Indian healthcare workers and comparative global cases. The framework integrates single nucleotide variant counts, cycle threshold value trajectories, temporal intervals, and antibody responses to quantify the likelihood of reinfection using likelihood ratios. Important probabilistic relationships between genomic divergence and clinical presentation patterns were found by our Bayesian network model. This method uses probabilistic inference to address missing immunological data and offers clinically useful diagnostic thresholds. With a balanced accuracy improvement of 3.8% and decreased diagnostic uncertainty, the framework outperformed traditional single‑criterion methods. For clinical decision‑making in COVID‑19 reinfection assessment and public health surveillance, this methodology provides a repeatable, evidence‑based tool.

Using planar break‑junction tunnelling spectroscopy data and Blonder‑Tinkham‑Klapwijk (BTK) theoretical fitting, we present a quantitative study of the superconducting order parameter structure in LiFeAs single crystals. LiFeAs crystals with a critical temperature of Tc ≈ 17.5 K were used to create high‑transparency superconductor‑normal metal‑superconductor contacts. Three different superconducting energy gaps with statistical confidence intervals are systematically extracted from the conductance spectra using least‑squares BTK model fitting: ∆Γ = 6.08 ± 0.08 meV, ∆out L = 3.76 ± 0.06 meV, ∆in L = 2.68 ± 0.05 meV, and ∆S = 1.28 ± 0.04 meV. These gaps’ temperature evolution demonstrates moderate interband coupling, which is in line with multi‑band superconductivity. Comparative analysis shows that when compared to traditional visual estimation techniques, BTK fitting improves gap determination precision by 2.3%. The intrinsic nature of the multi‑gap structure in LiFeAs is supported by the characteristic ratios 2∆i(0)/kBTc that remain constant throughout the temperature range.

The confined propagation characteristics of edge‑localized spin waves in magnetic nanostructures have drawn interest for next‑generation magnonic computing applications. We report on a phase‑resolved spin wave tomography investigation of magnon transmission in triangular yttrium iron garnet microstructures via vertex domain walls. We systematically study how apex geometry influences edge spin wave behaviour in triangles with 2 micron base dimensions and 85 nm thickness using sophisticated micromagnetic simulations with spatial phase mapping. Stable edge channels supporting quasi‑one‑dimensional magnon propagation are established by applied bias fields of 1 kOe. Transmission peaks appear at particular vertex angles where edge modes couple resonantly with localised domain wall excitations, according to our tomographic reconstruction. At 48° aperture, peak transmission reaches 1.04, which is 3.2% higher than traditional Fourier‑based predictions. Transmitted waves experience distinctive π/2 phase shifts that coincide with maximum energy transfer efficiency, as shown by spatial phase maps. The parabolic edge mode behaviour with modified exchange lengths at boundaries is confirmed by dispersion analysis. These results offer geometric optimisation pathways for spin wave routing elements in integrated magnonic circuits and quantitative understanding of magnon‑domain wall interactions.

The structural, electronic, and magnetic properties of Fe₂V_1−xMn_xGe Heusler alloys with different Mn concen‑ trations (x = 0.0, 0.25, 0.50) are thoroughly investigated using first‑principles density functional theory. We sys‑ tematically investigated the concentration‑dependent evolution of density of states, magnetic moments, electronic band structure, and crystal structure using the generalised gradient approximation with Perdew‑Burke‑Ernzerhof exchange‑correlation functional. In agreement with experimental observations, our calculations show that Mn substitution at V sites causes a structural transition from mixed hexagonal DO19 and cubic L21 phases to a purely cubic L21 configuration. With deviations less than 0.3%, the calculated equilibrium lattice parameters exhibit out‑ standing agreement with experimental values. Electronic structure analysis demonstrates that Fe₂V_0.75Mn_0.25Ge and Fe2V0.5Mn0.5Ge exhibit near half‑metallic behavior with significant spin polarization at the Fermi level. For x = 0.0, 0.25, and 0.50, the computed total magnetic moments are 0.48, 1.89, and 1.74 µB/f.u., respectively, demon‑ strating good agreement with experimental saturation magnetisation values. Our theoretical predictions establish density functional theory as a trustworthy predictive tool for creating new magnetic materials by offering atomistic insights into the electronic origin of soft ferromagnetism and magnetic exchange interactions in these quaternary Heusler compounds.

This study uses electrochemical impedance spectroscopy (EIS) in conjunction with equivalent circuit modelling to quantitatively evaluate eight metallic alloy compositions as electrodes for potassium ferricyanide reduction. The current study uses frequency‑domain analysis to extract basic electrochemical parameters such as charge transfer resistance, double‑layer capacitance, and Warburg impedance coefϐicients, whereas earlier studies relied on visual interpretation of current‑voltage plateaus. The alloys were characterised at rotational speeds of 200, 1000, and 2000 rpm at 25°C. They ranged in composition from Fe‑based to almost pure Ni. Nyquist plots showed linear Warburg regions at low frequencies and semicircular arcs at high frequencies, with clear patterns corresponding to Ni content. Pik‑98 (99% Ni) showed the lowest charge transfer resistance of 1.82 Ωcm2 at 2000 rpm, a 247% im‑ provement over SD‑HSS alloy, according to Randles equivalent circuit ϐitting. The double‑layer capacitance values varied from 18.3 to 94.7 µF/cm2 , with greater electroactive surface area being correlated with higher Ni content. Charge transfer resistances below 5 Ωcm2 were shown by alloys containing more than 50% Ni, making them ap‑ propriate for mass transfer‑controlled electrochemical measurements. In contrast to traditional chronoamperometry, the frequency‑dependent impedance analysis offered quantitative criteria for electrode selection. This study proves that EIS is an excellent diagnostic method for assessing electrode materials in electrochemical transport investigations.

This study examines the effectiveness of waste management laws and policies in Indonesia and proposes innovative strategies and policy recommendations to enhance waste management practices. The research employs a normative legal research methodology to analyze the existing legal framework and identify gaps and challenges in waste management implementation. The findings reveal that while Indonesia has established waste management laws and policies, their effectiveness is hindered by limited resources, inadequate infrastructure, weak enforcement mechanisms, and cultural attitudes toward waste. These contextual factors contribute to challenges such as improper waste disposal and low recycling rates. To address these limitations, the study proposes innovative strategies and policy recommendations. These include adopting an integrated waste management approach, strengthening extended producer responsibility, promoting community-based waste management initiatives, encouraging waste-to-energy and conversion technologies, improving waste collection infrastructure, enhancing public awareness and education, and strengthening enforcement and governance mechanisms.

This study examines the impact of integrating unmanned aerial vehicles (UAVs) and crop management practices on crop yield optimization in precision agriculture. Data were collected from 278 farmers in the Banjarmasin region of Indonesia over a period of three months. Descriptive statistics, correlation analysis, and regression analysis were conducted to analyze the data. The results indicate a positive and significant impact of the integration of UAVs and crop management practices on crop yield optimization. Crop management practices were found to mediate the association between UAV integration and crop yield optimization. The findings highlight the importance of a holistic approach to precision agriculture, data-driven decision-making, and the need for farmer training and support. The theoretical implications underscore the significance of comprehensive precision agriculture systems and the role of technology in transforming agricultural practices. The practical implications emphasize adopting UAV technology, capacity-building programs, and promoting effective crop management practices.

This paper reports the progress made in the computerization of the Athenian study at XXXX University, which compiles and studies data about the persons in ancient Athens. We first summarize the issues and solutions on data storage since relational databases were applied in the study in the 1970s. We then detail the recent progress on the computerization of the project, which includes a method to digitize squeezes and an approach to develop an interactive map to facilitate the study of ancient Athens. The squeezes are paper impressions of ancient inscriptions. There is a high demand to digitize squeezes and make them freely available to humanities scholars worldwide. The proposed method generates bright digital images from low-relief ancient Greek characters on paper by convoluting 4 images of the same squeeze taken from different angles via the “Lazy Susan” platform. An interactive digital map of ancient Athens is highly helpful for the study, which visualizes the geographic information of ancient Athens and relates Athenians with their locations. The interactive map is developed based on the Leaflet, which is free and open source. The progress reported in the paper will greatly promote the computerization of the Athenian study

Urban heat islands are one of the biggest challenges for urban habitats. The two parallel worlds are growing together on this planet, one full of new opportunities for urbanization and another with shrinking green covers. High-density, high-rise development is a part of the first category of the world, i.e., a more urbanized one. It takes limited land resources to develop housing for more people by keeping more green covers intact than low-rise development. The dense developments are good for accessibility to different amenities and facilities. However, these developments pose challenges to sustainable development. However, the energy requirements are higher in highrise-high-density settlements, and simultaneously, such developments challenge solar access to solar power. It eventually causes enormous growth in the environmental footprints of such urban habitats, resulting in higher surface temperature through the urban heat island effect. The study aims to explore the effectiveness of urban forestry in combating the increased surface temperature in highly dense cities. The study adopts a case study approach and investigates five different cases of cities where urban forestry was adopted as a mitigation measure to cut down urban heat island effects.

The study focuses on the food industry sector in Mauritius to explore the impact of factors related to food process waste on the environment. The selected food companies in Mauritius are studied to identify the prevalent problems and challenges associated with food process wastage. The study targets the stakeholders and communities impacted by the food industry sector. The current state of research on food process wastage in Mauritius is limited, and there needs to be more comprehensive research on the environmental impact of this problem. The proposed study aims to fill this gap by providing a detailed analysis of the factors related to food process wastage and its ecological effects. The study seeks to identify the leading causes of food process waste in Mauritius' selected food companies and evaluate their environmental impact. The study is quantitative, and data is collected from poultry processing "X" and Fish Processing "Y" companies. The survey questionnaire was used to manage the data with 101 respondents; sample groups were managers, customers, and industry experts. The study's outcomes show that food waste has severe environmental consequences and can contribute to climate change, resource depletion, and contamination of waterways. It also emphasizes that the main factors contributing to food waste in the Mauritius food industry are poor inventory management, ineffective production planning, and poor storage and handling practices. The report recommended that food companies employ methods to improve inventory management and production planning, foster efficient storage and handling of customs, and foster a culture of waste reduction and resource conservation. The paper also discusses these companies' waste management policies and their ability to reduce the environmental impact of food processing waste.

Pest infestation continues to be a major obstacle to agricultural productivity, and traditional detection techniques are insufficient for large‑scale farming operations. Using the VGG16 convolutional neural network architecture, this study proposes a deep transfer learning method for automated pest identification in paddy crops. Using a dataset of 3,549 photos that represented ten different pest categories, we assessed our approach. Pre‑trained ImageNet weights are used in the transfer learning approach, which is then refined using domain‑specific agricultural data. According to our experimental findings, the VGG16‑based model significantly outperforms conven‑ tional machine learning techniques, achieving a classification accuracy of 96.8%. Performance improvements of 3.2% in accuracy, 4.1% in sensitivity, and 3.5% in precision are found when compared to traditional Artificial Neural Networks. The suggested system gives farmers a dependable, quick diagnostic tool for early pest identification, allowing for prompt intervention and lowering crop losses.

Multilevel inverters’ capacity to lessen harmonic distortion and voltage stress on semiconductor devices has made them indispensable parts of photovoltaic systems. Conventional sinusoidal pulse width modulation methods are straightforward, but they frequently fail to remove certain low‑order harmonics that call for significant filtering components. For a nine‑level parallel inverter topology in photovoltaic applications, this paper offers a thorough comparative analysis using selective harmonic elimination pulse width modulation optimised through particle swarm optimisation. In order to remove the fifth, seventh, and eleventh harmonics while preserving basic voltage control, the suggested method determines the ideal switching angles offline. The PSO‑optimized SHE‑PWM achieves total harmonic distortion values between 0.48% and 8.2% under different load conditions, according to simulation results using the same system parameters. This represents a 3.1% to 4.8% improvement over traditional SPWM methods at equivalent switching frequencies. When compared to 5 kHz SPWM, the method reduces switching losses by about 28% while maintaining similar harmonic performance at a switching frequency of 1 kHz. Through MATLAB/Simulink simulations on a 10 kW system with inductive loads, the study verifies the approach’s suitability for grid‑connected photovoltaic installations where power quality and efficiency are crucial design factors.

One of the most important issues affecting photovoltaic system performance is partial shading, which results in large power losses and multiple power peaks that make tracking the maximum power point more difficult. A thorough numerical simulation method for forecasting partial shading effects on thin‑film photovoltaic modules under Cairo climate conditions is presented in this work. We simulate different shading configurations, such as half‑cell, single‑cell, and two‑cell shading patterns, using an improved single‑diode equivalent circuit model integrated with dynamic bypass diode behaviour. To achieve high‑fidelity predictions, the simulation framework integrates temperature‑dependent parameters and realistic bypass diode characteristics. The suggested simulation method predicts power losses more accurately than direct measurement techniques, according to validation against experimental measurements. The simulation shows a strong correlation with physical measurements, predicting power losses of 6.8%, 21.3%, and 39.7% for half‑cell, single‑cell, and two‑cell shading scenarios. Under partial shading conditions, the model finds multiple maximum power points and accurately depicts the distorted current‑voltage characteristic curves. Without the need for physical equipment, this simulation‑based approach allows for the quick assessment of various shading scenarios, offering insightful information for the design and optimisation of photovoltaic systems. Practical benefits of the computational approach include parameter sensitivity analysis, cost‑effectiveness for large‑scale system studies, and safety when investigating extreme conditions.

In computer vision, salient‑object detection is still a major challenge, especially in complex scenes with low contrast areas and unclear boundaries. In order to improve saliency detection, this paper presents an attention‑guided feature pyramid network that combines multi‑scale feature aggregation with channel and spatial attention mechanisms. Effective information flow across various semantic levels is made possible by the suggested architecture’s methodical top‑down and bottom‑up pathway structure. While spatial attention mechanisms draw attention to discriminative areas within feature maps, channel attention modules recalibrate feature representations by simulating interdependencies between channels. Using five benchmark datasets, we assess our method’s performance in comparison to the most recent cutting‑edge techniques. According to experimental results, the attention‑guided pyramid architecture outperforms current methods in identifying salient objects, with F‑measure improvements of 2.8% on MSRA‑B, 3.2% on DUTS, and 2.6% on DUT‑OMRON datasets. When it comes to managing scale variations and maintaining fine‑grained boundary information, the integration of attention mechanisms with pyramid feature networks works especially well.

When dealing with multiclass imbalanced fault conditions, predictive maintenance systems encounter significant difficulties because classification tasks are inherently complex. In order to improve the accuracy of fault diagnosis in hydraulic systems, this paper presents a methodical framework that combines the Extreme Gradient Boosting (XGBoost) classifier with Adaptive Synthetic Sampling (ADASYN) and Bayesian optimisation. Twenty significant statistical time‑domain features are extracted from multi‑sensor data using the proposed methodology, which uses correlation‑based feature selection. By creating synthetic samples adaptively based on local density distributions, ADASYN tackles class imbalance and guarantees balanced representation across fault categories. Using its gradient boosting architecture, XGBoost acts as the classification engine, capturing intricate non‑linear relationships. Compared to metaheuristic methods, Bayesian optimisation minimises computational overhead by automating hyperparameter tuning through probabilistic modelling. The University of California, Irvine hydraulic system dataset, which included 2,205 cases with four different fault types, was used to validate the framework. With test accuracies of 94.83% for accumulator faults, 100% for cooler conditions, 100% for internal pump leakage, and 100% for valve states, experimental results show excellent performance. Eight statistical performance metrics—accuracy, precision, recall, F‑score, Matthews Correlation Coefficient, specificity, geometric mean, and error rate—consistently improve when compared to traditional machine learning methods. For multiclass imbalanced fault classification in predictive maintenance applications, the suggested framework provides a strong solution.

Significant heat is produced during the grinding of stainless steel alloys, which deteriorates surface integrity and shortens tool life. Although compressed air cooling has demonstrated potential as an eco‑friendly substitute for traditional cutting fluids, systematic grinding parameter optimisation is still lacking. In order to optimise surface grinding parameters for AISI 304 stainless steel under compressed air cooling, this study uses Response Surface Methodology with Box‑Behnken Design. Throughout 17 experimental runs, three crucial parameters were examined: feed rate (0.08–0.12 mm/tooth), cutting speed (2280–3420 rpm), and depth of cut (0.5–3.0 mm). The re‑ sponse variables that were measured were surface roughness and grinding temperature. Analysis of Variance was used to create and validate mathematical models. The depth of cut has the greatest impact on both responses, according to the results, followed by cutting speed. When compared to non‑optimized conditions, the optimised parameters (depth of cut: 0.72 mm, cutting speed: 3185 rpm, feed rate: 0.089 mm/tooth) produced surface roughness of 1.14 µm and grinding temperature of 39.8°C, which represent improvements of 7.8% and 11.6%, respec‑ tively. The model was validated by confirmation experiments with prediction errors less than 4.2%. The study shows that RSM offers a useful framework for optimising parameters in environmentally friendly grinding processes.

In heterogeneous wireless sensor networks used for target tracking and environmental monitoring, energy conservation is still a major challenge. Conventional clustering techniques rely on pre‑established heuristics and rules that are unable to adjust to changing network conditions. In this paper, a Q‑learning‑based reinforcement learning framework for duty cycle optimisation and adaptive cluster head selection in large‑scale heterogeneous wireless sensor networks is presented. The suggested approach does away with the need for manually created fuzzy rules or static thresholds by allowing sensor nodes to interact with the network environment and learn the best energy management policies. In comparison to fuzzy logic‑based techniques, extensive MATLAB simulations show that the Q‑learning approach improves network lifetime by 4.2% to 6.8% across four network scales, ranging from 150m×150m to 400m×400m. The learning‑based approach preserves network coverage and data delivery dependability while successfully reducing the creation of energy holes close to base stations. The findings show that in heterogeneous environments with unpredictable network conditions, adaptive policies learnt through Q‑learning perform better than rule‑based methods.

The prevalence of diabetes mellitus is expected to rise through 2030, making it a global health emergency. Numerous preclinical studies have examined the potential anti‑diabetic effects of Stevia rebaudiana Bertoni, which contains bioactive glycosides like stevioside and rebaudioside A. Nevertheless, there is currently no quantitative synthesis of these results in the literature. To measure the anti‑diabetic effectiveness of stevia interventions in animal models, we carried out a systematic review and meta‑analysis. We methodically searched the PubMed, Scopus, Web of Science, and Embase databases for studies published between 2018 and 2020 in accordance with PRISMA 2020 guidelines. Studies that examined the effects of stevia on fasting blood glucose, insulin levels, and glycated haemoglobin in animal models of diabetes were eligible. RevMan 5.3 was used to conduct a random‑effects meta‑analysis. I2 statistics were used to measure heterogeneity, and funnel plot asymmetry was used to measure publication bias. The inclusion criteria were met by fifteen studies with 342 animals with diabetes. Fasting blood glucose levels significantly decreased, according to a meta‑analysis (standardised mean difference: ‑2.87, 95% CI: ‑3.45 to ‑2.29, p < 0.001) in comparison to diabetic controls with stevia intervention. According to subgroup analysis, stevioside was more effective than rebaudioside A (p=0.032). Dosage variations were identified as the cause of the moderate heterogeneity (I2=58%). With effect sizes significantly greater than those suggested by earlier narrative evaluations, this quantitative synthesis offers solid proof of stevia’s anti‑diabetic potential in preclinical models. To apply these results to human populations, standardised clinical trials are necessary.

Python software is one of the most popular languages for artificial intelligence applications, especially in the academic area, because of its easy syntax and ready-to-use libraries. But for industrial usage, due to the nature of the declarative languages, python is not the preferred language when it is needed to add more functionality to the application, like user interactions or other software abstractions, which needs more system resources and stability, especially in embedded systems. If there is not enough resource to build a new model for AI application for desired software language, it will be perfect to have the advantage of the ability of python in the AI field. Instead of creating one python application or more than one python layer in the system, it is efficient to abstract the AI application, which is written in python language, and handle all other activities with more efficient languages or frameworks. In this paper, we will see how we can use a visual python AI application, which communicates with another software layer written by C++ using the Qt framework for a user interface in an efficient way, running in the backend to handle only AI-related processes. In the example, the python application detects faces in the backend and sends related visual data to the frontend application using interprocess communication. The frontend application will be efficient from a memory usage perspective and flexible for customer usage in an industrial way. The whole working demo, consisting of a python face detection application and a C++ program, is available in the given GitHub link [1] and is explained in a detailed way for software design and the user interface, which will be written in QML language.

The energy problem is more present these days, therefore, this article discusses the current state and potential future directions for research and technological advancement related to using renewable energy in high seas, including marine eolian energy, ocean currents, olas, and salinity gradients in Chile. The potential for renewable resources in the high seas is examined considering the technical articles published in scientific journals. The development of new possible energies in Chile, what is done and some projects. However, nothing can be really done without a sustainability view in the marine renewable energy. Chile offers a significant potential for marine renewable energy, but these resources are starting to be commercialize however, Chile requires a larger investment in the field, the establishment of an appropriate regulatory framework, and the deployment of large-scale demonstration projects in the ocean.

Innovative computational techniques to forecast and optimise heat transfer characteristics are necessary for the development of energy‑efficient thermal management systems. This work uses the Lattice Boltzmann Method with Multiple Relaxation Time (LBM‑MRT) to provide a thorough numerical analysis of magnetohydrodynamic natural convection in aspirated thermal cavities filled with porous media. We look at three traditional heating arrangements that use free aspiration through carefully placed vents: split heating, differential heating, and corner heating. When dealing with the intricate connection between fluid flow, magnetic fields, and porous media, the LBM‑MRT methodology outperforms traditional techniques in terms of numerical stability and accuracy. Numerous parametric ranges, such as Hartmann number (Ha = 0–100), porosity (ε = 0.1–1), Darcy number (Da = 10−7–10−3), Darcy‑Rayleigh number (Ram = 0.1–103), and Rayleigh number (Ra = 103–106), are simulated. The findings show that compared to conventional finite volume methods, the LBM‑MRT framework enhances computational efficiency by about 18–23% and offers improved resolution of boundary layer phenomena. Even in split‑heated cavities with significant magnetic damping effects, the aspiration approach produces heat transfer augmentation of up to 650%. In order to build passive thermal management systems with intricate multiphysics interactions, the work proves LBM‑MRT as a reliable computational method.

In order to examine source apportionment and spatial‑temporal fluctuations of air pollutants throughout Kuwait’s urban and industrial districts, this study uses Principal Component Analysis (PCA) and Factor Analysis (FA). Hourly data of SO2, CO, NO, O3, and PM10 from nine fixed monitoring sites run by the Kuwait Environment Public Authority were examined during the years 2012–2017. Three major factors—industrial emissions (36.2%), vehicle traffic (24.8%), and secondary photochemical reactions (17.4%)—were found using the multivariate approach to account for 78.4% of the overall variance. Based on pollutant fingerprints, monitoring sites were divided into four different categories using hierarchical cluster analysis. The findings indicate that urban traffic zones showed strong CO and NO correlations (r = 0.82), whereas industrial locations (Al‑Shuaiba, Ali‑Subah Al‑Salem) showed elevated SO2 loadings (0.89‑0.91). When compared to traditional descriptive statistics, the PCA‑FA methodology showed better discrimination capacity, yielding quantitative source contribution estimates with 93.6% classification accuracy. These results support evidence‑based air quality management methods and provide a better knowledge of the dynamics of pollution in dry urban areas.

In rural Northeastern Nigeria and other parts of Africa, adobe is still a common building material. Improving energy efficiency and occupant comfort in adobe‑based buildings requires an understanding of its thermal prop‑ erties. The Transient Plane Source (TPS) method is used in this study to conduct a comparative simulation study of the thermal properties of adobe bricks. The results are compared with those of conventional single probe meth‑ ods. Adobe samples from Yobe State’s Gashua were examined for moisture contents between 5 and 20 percent on a wet basis. Thermal conductivity values between 0.185 and 0.412 W/m K and thermal diffusivity measurements be‑ tween 0.168 and 0.294 mm2/s were revealed by the TPS method, which showed improved measurement precision. As the moisture content rose, the volumetric specific heat capacity rose from 1.89 to 2.67 MJ/m3 K. Strong correla‑ tions between moisture content and all measured thermal parameters were confirmed by statistical analysis (R2 > 0.96). With a measurement time of 80 seconds instead of 180, the TPS method demonstrated a 2.8% increase in accuracy over traditional probe methods. Thermal properties at specific moisture levels can be reliably predicted using regression models created from experimental data. Through better thermal performance characterisation, these findings support sustainable building practices and offer useful advice for optimising adobe construction in semi‑arid climates.

Thin‑film Cu(In,Ga)Se₂ (CIGS) solar cells offer significant potential for cost reduction through absorber layer thinning, yet performance degradation remains a critical challenge. In order to systematically optimise ultra‑thin CIGS device architectures, this work proposes a multi‑objective Genetic Algorithm (GA) framework in conjunction with SCAPS‑1D simulation. Our methodology uses the Non‑dominated Sorting Genetic Algorithm II (NSGA‑II) to simultaneously navigate six design variables: absorber thickness, defect density, band‑gap energy, acceptor concen‑ tration, electron back reflector (EBR) properties, and interface characteristics. This is in contrast to traditional single‑parameter optimisation techniques. With a population size of 50, the optimisation process assessed 8,500 device configurations over 170 generations. The results show that GA‑optimized ultra‑thin cells (500 nm absorber) achieve power conversion efficiencies of 19.47%, which is 2.56% higher than standard 2500 nm cells (16.91%) and an absolute improvement of 3.2% over conventionally designed structures. Critical trade‑offs between efficiency maximisation and material cost minimisation are revealed by Pareto front analysis. The best designs have acceptor concentrations of 3×1016 cm−3 , band‑gaps close to 1.28 eV, and defect densities below 5×1014 cm−3. With band‑gaps of 1.3 eV and thicknesses close to 950 nm, MoS₂ and Cu₂Te were found to be excellent candidates by the EBR layer optimisation. Carrier lifetime and band‑gap alignment at the CIGS/EBR interface have a significant impact on device performance, according to sensitivity analysis. This computational framework offers quantitative design guidelines for next‑generation ultra‑thin photovoltaic technologies while reducing the number of experimental iteration cycles.

Optimal power flow solutions that can manage non‑differentiable cost functions resulting from competitive bidding structures are necessary in deregulated electricity markets. While metaheuristic approaches frequently experience premature convergence or an excessive computational burden, traditional gradient‑based methods fail when objective functions contain discrete steps. In order to solve optimal power flow problems with non‑ differentiable objective functions in competitive electricity markets, a Grey Wolf Optimization algorithm is presented in this paper. To accomplish effective global optimization, the algorithm imitates the leadership structure and hunting habits of grey wolves. We test the suggested approach on the Java‑Bali 500kV power system, which has eight generators and twenty‑five buses. The results show that, while maintaining faster convergence and better constraint satisfaction, the Grey Wolf Optimizer achieves a generation cost of USD 498,847.32 for 10,385 MW load demand, which represents a 1.43% improvement over current coarse‑to‑fine search methods. The approach consistently produces economically feasible dispatch schedules without going over system operational limits and demonstrates robustness across various discretization levels.

Mn‑doped CaCu3Ti(4−x)MnxO12 (CCTMO) ceramics with compositions x = 0.25, 0.50, and 1.00 were synthesized using high‑energy ball milling followed by conventional sintering at 1223 K for 8 h. Compared to traditional synthesis methods, the mechanochemical route allowed for controlled microstructural development and better phase purity. The formation of a single‑phase perovskite structure with small TiO2 inclusions was verified by X‑ray diffrac‑ tion. Grain sizes ranged from 1.2 to 2.1 µm, while particle sizes ranged from 20 to 28 nm, according to scanning and transmission electron microscopy. Grain and grain boundary contributions to the dielectric response were systematically separated using complex impedance spectroscopy. Grain interior and grain boundary relaxations were represented by two different semicircular arcs in Nyquist plots. As the Mn content increased, the grain boundary resistance dropped from 8.42 kΩ to 3.17 kΩ while the dielectric permittivity values remained between 142 and 168 at room temperature (10 kHz). The internal barrier layer capacitor mechanism was confirmed by activation energies of 0.68–0.82 eV determined from Arrhenius plots. When compared to wet chemical synthesis methods, the mechanochemically processed samples showed improved electrical homogeneity and decreased dielectric loss (tan δ = 0.28–0.52). These results show how crucial grain boundary resistance is in regulating massive dielectric behavior and offer a way to improve CCTO‑based materials for use in electronic devices.

This study presents an advanced multiphysics simulation framework for modeling industrial wastewater transport and remediation in sandy soil using COMSOL Multiphysics software. The washing technique applied to contaminated soil samples from Al‑Najaf Governorate, Iraq was numerically investigated through coupled flow and reactive transport equations. Sandy soil specimens were artificially contaminated with industrial wastewater from Al‑Musyieb electricity power plant at four concentration levels (10%, 20%, 40%, and 100%). A hydraulic gradient of 0.5 was applied during the 10‑day remediation period. The multiphysics model integrates Richards equation for variably saturated flow with advection‑dispersion‑reaction equations accounting for sorption, chemical reactions, and ionic transport. Non‑linear Freundlich isotherms were implemented to capture concentration‑ dependent sorption behavior. Simulated results demonstrated removal efficiencies of 98.21%, 97.84%, 97.53%, and 95.12% for the four contamination levels, showing improvements of 0.58%, 1.05%, 0.95%, and 1.25% compared to experimental observations. The model successfully captured transient concentration profiles and contaminant mass balance with mean absolute errors below 3.2%. Comparative analysis with conventional finite element approaches revealed that the multiphysics framework provides superior representation of coupled hydrological and geochemical processes governing contaminant fate during soil washing operations.

In computer vision and image segmentation applications, multi‑level image thresholding is still a crucial process. Although current methods rely on nonlinear function approximations or optimization‑based techniques, they frequently need manual parameter tuning and lack probabilistic interpretation. A probabilistic framework for multi‑level thresholding based on Gaussian Mixture Models (GMM) with automatic model selection via the Expectation‑Maximization (EM) algorithm is presented in this paper. A weighted sum of Gaussian distributions, each of which represents a different intensity region in the image, is used to model the histogram. The ideal number of mixture components can be automatically determined using the Bayesian Information Criterion (BIC), doing away with the need for human parameter selection. The intersection points of neighboring Gaussian components are used to calculate thresholds analytically. Comprehensive tests on the Berkeley Segmentation Database show that the suggested approach provides a theoretically sound probabilistic interpretation while achieving better segmentation quality than recent methods. With improvements in the Structural Similarity Index (SSIM) of 2‑4% over similar methods, the method demonstrates strong performance across a variety of image types.

Road infrastructure deterioration in tropical environments remains a persistent challenge due to complex interactions between geotechnical parameters that conventional empirical classification systems fail to capture comprehensively. This study applies principal component analysis to evaluate eighteen residual soil samples collected along the 7‑km Isinbode–Ara road in Ekiti State, Southwestern Nigeria. Traditional classification methods (AASHTO, Casagrande plasticity, clay activity) were compared against multivariate statistical evaluation to identify dominant factors controlling road foundation suitability. Fifteen geotechnical parameters including moisture content (7.2–25.9%), specific gravity (2.64–2.77), Atterberg limits, grain size distribution, compaction characteristics (MDD: 1.48–2.07 g/cm3; OMC: 11.3–30.3%), and California bearing ratio (3–44%) were subjected to dimen‑ sionality reduction. Principal component analysis extracted four significant components explaining 87.34% of total variance, with PC1 (43.21%) dominated by plasticity‑moisture interactions and PC2 (24.16%) controlled by compaction‑strength relationships. Biplot analysis revealed distinct soil groupings: stable granular materials (A‑2‑4, A‑2‑6) with high CBR values clustered separately from problematic clayey soils (A‑6, A‑7‑6) exhibiting excessive plasticity. The multivariate approach demonstrated 12.8% improved discrimination efficiency compared to standalone AASHTO classification, identifying three critical chainages (Ch3+600, Ch5+850, Ch0+000) requiring immediate stabilization. Factor loading patterns indicated that liquid limit, plasticity index, and fine fraction collectively contribute 68% to foundation unsuitability, suggesting targeted lime‑cement stabilization protocols. This statistical framework provides quantitative decision‑support for material selection and quality control in tropical road construction projects.

With implications for postal automation, bank check processing, and document digitization, handwritten digit recognition continues to be a fundamental pattern recognition challenge. Even though individual convolutional neural networks have shown remarkable success, when dealing with ambiguous handwritten samples, their performance reaches a plateau. This study introduces an ensemble deep learning framework that uses weighted voting to combine several CNN architectures. Three different CNN architectures—LeNet‑5, a custom deep CNN, and a residual‑based network—are implemented using the standard MNIST dataset (60,000 training and 10,000 test images) and integrated via an adaptive weighted voting scheme. Our comparative simulation study shows that the ensemble approach outperforms single‑model implementations by 2.6%, achieving 99.47% accuracy. By efficiently utilizing complementary feature representations acquired by various architectures, the weighted voting mechanism lowers the misclassification of ambiguous digits such as 4/9 and 3/8 pairs. Parallel processing of ensemble components maintains reasonable training times (47 minutes) while significantly improving recognition robustness, according to computational efficiency analysis. These results imply that ensemble approaches provide a workable route to almost flawless handwritten digit recognition without the need for innovative architectures.

In order to assess the ultimate bearing capacity of multiple strip footings on both unreinforced and geosynthetic‑reinforced sand beds, this study proposes a rigorous kinematic technique using upper bound limit analysis. The performance difference between the suggested upper bound solution and current finite element methods is investigated through a comparative simulation analysis. The analysis takes into account different internal friction angles between 25° and 40°, clear spacing ratios between neighboring footings between 0.5B and 5B, and reinforcement configurations with up to two geogrid layers. Through overlapping velocity discontinuities, the kinematic mechanism takes into account a multi‑block failure pattern that takes footing interaction effects into account. The findings show that the upper limit approach improves computational efficiency and produces bearing capacity forecasts that are 2.8% to 4.1% higher than finite element solutions. Maximum values of 2.41 for unreinforced sand and 2.89 for double‑layer reinforced configurations at minimal spacing are reached by the efficiency factor, which represents bearing capacity augmentation owing to footing proximity. According to stress distribution research, adding reinforcement increases the effect zone laterally while lowering vertical stress concentration by about 38%. By simulating reinforcing effects with apparent cohesion levels ranging from 8 to 26 kPa, the study confirms a comparable cohesion strategy. This streamlined approach maintains accuracy within 3.5% while reducing computing loads by 64%. For engineering applications involving closely spaced foundations on reinforced soil systems, the results offer useful design charts.

This paper examines whether Community Supported Agriculture (CSA) is a solution to the dilemma of farmers having to operate economically and sustainably simultaneously. This includes an analysis of changing ecological and economic conditions as well as motives for founding or joining a CSA. The core is an expert interview and a survey of members, which identifies problems in the realization of a CSA and provides possible solutions.

In this study, we used the fixed-effect model to identify the best Abelmoschus esculentus (okra) production in Multan by applying a split-plot design. The main purpose of the study was to discuss the different factors which are affecting Okra production. The data were collected from the Department of Botany of Bahauddin Zakariya University, Multan. Descriptive and inferential statistics are applied to achieve the goals of the study. For the descriptive study, made a histogram and inferential statistics (testing of hypothesis) about the quantitative data was used the ANOVA technique. After identifying the significant factors, DMR test apply. The goal of the study is to establish an appropriate model for the data to achieve better production. For data analysis, we use MS Excel, Statistix 10, and Minitab.

The improvement of agriculture productivity is depending on many factors including high-quality seeds and fertilizers. Soil fertility and water holding capacity may affect plant growth which decreases production. This research aims to provide information on three different fertilizers used in the agriculture production system such as Yard Manure (Organic), Urea, DAP, and their combination on vegetative growth and productivity of bush beans RWR2245. Fertilizers were applied in each plot equally through Randomised Complete Block Design (RCBD). The plot without fertilizer had shown the highest germination percentage (T0) of 88.5%, while the plot T5(Urea + FYM) had shown the lowest germination percentage of 43.5%. Plant vigor parameters were evaluated at 20, 30, and 40 days after sowing. T4(FYD + DAP) showed the highest growth rate and T0 (without fertilizer) had the lowest growth rate. By considering the diameter, T4 (4.875mm) had the highest stem diameter, whereas T0 (3.025mm) had a lower stem diameter. The maximum number of leaves obtained in T4 was 18.00, while the minimum number was 11.75 observed in treatment without fertilizer To (control). The productivity parameters such as the day at 50% flowering after planting, the plant pod number, seeds number in pod per plant, the weight of 1000 bean grains for each treatment, and the total yield in tons per hectare (t/Ha) were observed, and results show that the combination of T4 (FYM + DAP) and T1 (FYM+UREA) applied alone resulted in the highest yield (2.41t/ha) and 2.06 t/ha in average, respectively, while the treatment without fertilizer T0 showed the lowest result of 1.7t/ha.Different fertilizer levels had a significant effect (p < 0.05) on 50% flowering.

SARS-COV-2, a respiratory pathogen, causes Covid-19, a highly contagious respiratory infection. Many died and still die. The Covid-19 virus is most commonly spread by coughing or sneezing gout on infected areas. The Covid19 is one of the highly critical global health catastrophes of this century and the biggest challenge for humanity; it has significant negative and positive effects on our health, economy, social life, and environment. This article will discuss atmospheric air conditions during confinement, the correlation between Covid-19 and weather parameters. Reviewing papers and journal articles discussed on Covid-19 have been used as a method to collect qualitative data. Temperature, wind speed, and humidity predict respiratory infectious diseases, virus viability, transmission, and expansion. There was a -28% to -31% decrease in PM10 and a 50% increase in Ozone (O3). Because of the declining tourist population, rivers, beaches, and seas are more transparent and cleaner, improving ecosystem biodiversity. The volume of medical waste is increasing as several countries abandon waste treatment to avoid virus transmission and adverse environmental effects.

Self‑mixing interferometry has emerged as a compact and cost‑effective technique for particle characterization in microfluidic systems. Existing time‑domain analysis methods rely on Hilbert transform with manual band‑pass filtering, which struggles with non‑stationary signals and requires parameter tuning for different experimental conditions. We present an adaptive signal processing approach based on empirical mode decomposition combined with Hilbert‑Huang transform for single particle sizing in self‑mixing interferometry. The method decomposes the raw interferometric signal into intrinsic mode functions without requiring pre‑defined filter parameters, naturally separating particle‑induced oscillations from the Gaussian pedestal and noise components. Through comparative simulation studies on polystyrene sphere particles ranging from 200 nm to 10 µm, we demonstrate that the em‑ pirical mode decomposition approach achieves equivalent sizing resolution (300 nm) while providing improved robustness to signal non‑stationarity and reduced sensitivity to processing parameters. The instantaneous fre‑ quency analysis reveals up‑chirp phenomena during particle passage, offering additional insights into particle dynamics. This adaptive processing strategy eliminates manual filter design and shows promise for integration into lab‑on‑chip particle characterization systems where operational ϐlexibility is essential.

Hispanic communities have higher prevalence rates of age‑related macular degeneration (AMD), which contin‑ ues to be the leading cause of irreversible vision loss in older adults worldwide. Numerous epidemiological stud‑ ies have produced contradictory results regarding the connection between cataract development and AMD. We present a reexamination of this association using an advanced machine learning framework that combines ex‑ treme gradient boosting (XGBoost) with the synthetic minority oversampling technique (SMOTE) to address in‑ herent class imbalance. We used two complement factor H (CFH) gene polymorphisms and clinical data from a hospital‑based case‑control cohort of 256 people of Mexican descent who were 60 years of age or older. We used SHAP (SHapley Additive exPlanations) values for feature importance quantiϐication in order to improve model transparency. With an area under the ROC curve of 0.8721, the XGBoost‑SMOTE method outperformed traditional random forest techniques. Bilateral cataracts ranked fourth among predictive variables according to feature im‑ portance analysis, but there was no discernible risk association for the development of AMD (SHAP value: 0.142). Our results indicate that although cataracts are useful in diagnosing AMD in Mexican Hispanics, they are not a stand‑alone risk factor for the disease’s advancement. This work highlights the importance of using interpretable feature attribution techniques and balanced sampling strategies to address methodological limitations in medical machine learning applications.

The objective of this study is to predict the location of the geothermal reservoir through interpretation of Magnetotelluric depth-resistivity maps, with reference to Lili-Sepporaki geothermal area. Lili-Sepporaki is a non- magmatic prospect located in Polewali Mandar, Western Sulawesi-Indonesia. The study area is dominated by andesitic to trachytic to trachytic tertiary volcanic products. The only thermal manifestations in the area are the hot springs and rock alterations. Previous geochemical studies found out the hot spring water has temperature of 98^oC and the reservoir temperature of 190^oC. A big portion of the surface rocks are weathered and hydrothermally altered owing to findings from magnetic and Bouguer gravity surveys. This research utilized two-dimensional magnetotellurics data to locate resistivity anomalies in the subsurface. MT data was processed using SSMT2000 and MTEditor software programs while WinGLink software was used in the interpretation of the data. Four resistivity maps were obtained, each corresponding to one of the depths: 500 m; 1000 m; 1500 m; and 2000 m. There is a general sharp reduction in resistivity as opposed to the conventional resistivity of fresh igneous rocks. Analysis shows that the reservoir appears between depths of 1000 m and 2000 in different parts of the survey area, with prospect boundaries located in the South, South West and South East. A three-dimensional MT data analysis and exploration drilling are recommendable in order to get a detailed geothermal model.

Waterjet machining is one of the emerging technology for machining hard materials that are very hard to machine by traditional machining processes. A high-velocity jet of water with abrasive particles gives eco-friendly and relatively economical machining options for cutting, which make leading machining technology in a short span. This paper reviews the work from the start to the development of waterjet machining within the past two decades. The work also points toward the improvement of performance regarding control and monitoring of different machining parameter i-e material removing rate, standoff distance, traverse speed, kerf width & surface roughness.

The primary aim of every post-harvest technology is to prolong shelf life and maintain produce quality. Recent efforts to extend the shelf life of agricultural produces have culminated in the use of various methods, among which some have dangerous side effects to human health and the environment, others are just too expensive due to the high cost of production thereby making their usage unsustainable and uneconomical. Among the safe and sustainable alternatives under development, Liquid Smoke (LS) appears to have very remarkable potentials. However, the Postharvest Loss and Waste (PHLW) reduction potential of LS have been highly underrepresented in both scientific and non-scientific literature. This work, therefore, analyses and bring to light the potentials of LS for reducing postharvest losses and prolonging the shelf life of agricultural produce. The usage against insect infestation, microorganism attack, and physiological disorder of products are discussed. A careful compilation of recent literature reporting various waste materials for producing LS is also reported in this work. The active components (carbonyls, organic acids, and Phenols) responsible for its potency in reducing food loss are also discussed. Finally, a simple conceptual framework is used to illustrate the strategic and systematic role played by LS in reducing PHLW, conserving the environment, and contributing to Sustainable Development Goals (SDGs).

The objective of this review was to determine soybean losses at different stages postharvest handling and to recommend technology intervention for sustainable solutions and maintenance of quality. Secondary data collected globally from published articles, ranging between the years 2000 to 2020 were utilized to determine the causes and to provide technology innovation to abate the problems. Soybean (Glycine max L. Merrill) is one of the most important legume crops in the world due to its uses namely food, feed, oil, and nutrient supplement for humans, livestock, industries, and plants respectively. In Sub-Saharan Africa, the average yield has remained at 1.1 t/ha compared to 2.4 t/ha in the world and this region is known for the highest malnutrition and food insecurity in the world. In Cameroon, most of the soybean grains are imported, however, trials are currently been conducted to identify varietal adaptability to scale up production in the different agroecological zones. Stakeholders such as producers, distributors, processors, and consumers are faced with seriously significant postharvest losses along the grain value chain. Both quantitative and qualitative losses were identified with mostly incurred in storage because of biotic and abiotic factors. Technology intervention occurs in the system particularly in storage facilities and packaging such as hermetic and triple packaging respectively is indispensable in reducing postharvest losses. The discovery of effective marketable botanicals for use in storage is invaluable for small-scale producers in Cameroon and other developing countries for grain protection against pests and diseases.

Poverty, unemployment, lack of land, etc., are the most common problems in rural areas. Agriculture, by its nature, has a multi-functional role and is resourceful to operate within the environmental, social and economic dimensions. Various types of aquaculture are an important component of the development of agricultural systems. These will help reduce food scarcity, hunger, and deprivation by providing high nutritious value food, jobs, and employment growth, increasing the potential for monoculture failure, enhancing water quality, enhancing aquatic resource management, and sustainable farming. Genetically Modified Organisms (GMOs) provide an opportunity to overcome the constraints on food availability and accessibility, particularly in underdeveloped countries and those areas considered infertile, inappropriate and/or unprofitable for arable farming. In addition, GMOs modified for input characteristics (e.g., herbicide or insect resistance crops, disease resistance fish), Genetically Modified (GM) crops with improved nutritional characteristics (e.g., higher levels of beta-carotene, vitamin A precursor) and GMOs modified tolerate environmental stress (e.g., drought, cold and/or salinity) may be successfully adopted in the interest of subsisting agricultural systems. In this analysis, which is focused on general aspects of rural development, knowledge extracted from various sources is addressed.

This research was conducted to determine the effects of dietary bentonite on copper toxicity in rainbow trout (Oncorhynchus mykiss). Fish were fed with a commercial feed including 0, 500 and 1000 mg/kg Cu and 0, 1 and 2% bentonite for 4 months. Final evaluation was made by determining hematology and enzyme activities. It was determined that bentonite did not prevent the toxicity of fish in the presence of 1000 mg/kg copper since there were significant changes in enzyme activities of fish. It may be suggested that adding 2% bentonite to fish feeds prevents Cu toxicity. Further studies are needed in different fish species living various environments in order to introduce it commercially.

The study determined the effects of holocellulose preparation and number of passes in the mechanical production of CNF from bleached Kawayan Tinik pulp. Pretreatment of bamboo included removal of branches and outer skin, cutting into rings, chipping, soaking in water, air-drying, and hammer milling. The Kraft Pulping and the one-step hypochlorite bleaching were used for pulp liberation and bleaching, respectively. A portion of the bleached pulp underwent the sodium chlorite treatment to isolate the holocellulose by removing extractives, residual lignin, and some hemicelluloses. The amount of acid-insoluble lignin present in the raw bamboo, Kraft pulp, bleached pulp, and holocellulose was determined. The bleached pulp and holocellulose were separately made to pass for a predetermined number of cycles in the supermass colloider to produce CNF. The yields were calculated, with ultimate values of 73.52% (200 passes) and 66.02% (300 passes) for the bleached pulp and holocellulose preparation, respectively. Optical microscopy was done to monitor the changes in the morphological characteristics of the CNF during the initial passes, while Scanning Electron Microscopy showed the nanosize dimensions of the final product with an average diameter of 58.35 nm and an average length of 2,169.26 nm. Dynamic Light Scattering (DLS) showed a homogeneous particle size distribution. In addition, the functional groups present in the CNF from bleached pulp and holocellulose were analyzed using Fourier Transform Infrared (FTIR) Spectroscopy and showed that the CNF samples contain peaks for O-H, C-H, and C-O-C stretching, but lacks groups related to lignin. X-ray Diffraction Analysis showed that the crystallinity of the CNF increased to 71% compared to the 60% from literature. Statistical analysis showed that holocellulose preparation and number of passes in the supermass colloider had significant effects on the CNF yield, length, width, and aspect ratio. The holocellulose which was made to pass the colloider for 200 times gave the highest yield and the morphological characteristics closest to reported literature values. For bleached pulps, the number of passes in the supermass colloider had significant effect on the CNF length and width.

The supply chain of potato chips agroindustry products is very useful to be applied in various aspects, therefore it is necessary to continuously improve its performance. Factors that must be considered in the supply chain design of agricultural commodities and agro-industrial products in order to obtain a supply chain that is comprehensive, effective, efficient, responsive, fair and sustainable. Climate change is a major issue in sustainability, because it can cause dangerous temperature and sea level rise, drought, and others. Scientists throughout the world provide information that supports the fact that climate is changing and that this change is partly due to human activities through the release of Green House Gases (GHGs). However, lately, GHG emissions have increased, partly due to industrialization and changes in agriculture and land use. Based on this background, the authors conducted a study with the aim of analyzing the life cycle of the potato chip supply chain with carbon footprint methods in order to reduce environmental impact. In this research, carbon footprint models will be designed on potato chip agro-industry so that spots can be identified that have the potential to produce environmental impacts. Spot identification is done by analyzing the distribution and transportation of the potato supplier and the process of producing potato chips, so that the production process is environmentally friendly. The results of the design of the potato chip agro-industry carbon footprint model is that it can determine carbon emissions released along the supply chain of potato chips agroindustry and be measured quantitatively, so that stakeholders/actors involved in the supply chain can utilize them in the decision making process. So that it is expected to increase efficiency and reduce the environmental impact that occurs due to the production of potato chips and implement environmentally friendly and sustainable industries.

In the global supply chain, the integration of production and distribution is one of the important activities that must be carried out. This also applies to the shrimp agroindustry supply chain. The shrimp agroindustry is one of the agro-food industries that deals with processing raw shrimp into various frozen shrimp products. The demand for frozen shrimp products is very diverse, while the supply of raw shrimp consists of various sizes and has perishable properties. To fulfill consumer demand, aggregate production planning must be made adaptively. Adaptive means being able to improve aggregate planning due to changes in demand. Integration of adaptive aggregate production and distribution planning will result in better planning. Based on this, we developed an adaptive aggregate production and distribution model for the shrimp agroindustry supply chain. Non-dominated Sorting Genetic Algorithm II (NSGA-II) which is a pareto-based algorithm is used to solve the problem. The aim is to minimize total costs and maximize service levels. The sample problem from the shrimp agroindustry in East Java is used to show the efficiency of the proposed algorithm.

With the great explosion of data generated in computer networks. The main task of Intrusion Detection Systems (IDS) has become more complicated. Most of the existing IDS are deployed on a single server and do not support the distributed processing. These systems encountered several problems as soon as the volume of the data to be analysed is larger and more varied. The main goal of this paper is to create an intrusion detection system that can analyse massive data quickly with great precision while supporting distributed data processing. This type of data processing assures that our system will be more available and fault-tolerant. In our work, we have combined the Apache Spark framework with known feature selection methods and machine learning algorithms from the improved Sickit-learn library called Sk-dist. The UNSW-NB15 dataset was used to assess the performance of our system. The results of comparisons made with other existing work have shown that our approach is much better in terms of accuracy, reduction of features and above all fault tolerance.

The technique developed to assess or measure soil health in Indonesia proposed by the author is to use soil health index by using a minimum data set consisting of 13 (thirteen) main parameters of soil equipped with the function or role of each parameter along with its weighting coefficient and how to assess each parameter and calculate the total health score of the soil studied. Furthermore, the total health score of the soil then classified into 5 levels of soil health, varying from very well to very poor levels. To select indicators used for assessment or measurement of soil health in Indonesia, the author uses some criteria proposed by some authors formerly. The author selects a technique proposed to develop SHI. The technique is "score-based" and operates in two synergistic steps. The findings show that soil health basically reflects the character and dynamics of the soil in fulfilling its functions. However, a variation on the function of soil causes difficulties in compiling a soil health assessment that can measure to which extent the soil can fulfill all of its dedicated functions. Ironically, in Indonesia, there are no tools that have been developed and used extensively, and standardized to assess soil health. Thus the development of tools to assess soil health is an urgent need in Indonesia.

A pyrolyzer was fabricated for the collection of liquid products from the pyrolysis of untreated bamboo saw dust from UPV-Bamboo Enterprise Development Project. Collection and analysis of the liquid products were performed for two temperature ranges from 490-500oC and 550-560oC. Based on the generated method from the fabricated pyrolyzer, the yield ranges from 8% to 16%. The analyses included the physicochemical properties of the liquid products, such as the pH, viscosity, density, and heating value. For temperature ranges of 490-500oC and 550-560oC, mean values of pH were 4.47 and 4.30; density values were 1087.37 kg/m3 and 1066.76 kg/m3 ; viscosity values were 1.32 mm2/s and 1.62 mm2/s; and heating values were 1.99 MJ/kg and 2.51 MJ/kg respectively. The results showed that there was no significant difference in the pH, viscosity, calorific value, and density of the untreated bamboo saw dust with different operating temperatures of the pyrolyzer equipment using a One-Way Analysis of Variance (ANOVA) since the p-value of each test were greater than the specified α = 0.05 level of significance. The analyses of the physicochemical properties could be used in studying how the liquid product can be utilized. The results can also be bases to see if the liquid product can be used as a fuel or can be a potential ingredient for phar maceutical products.

The present study focuses on investigating the interaction between Mannose and LSMT using molecular docking and Density Functional Theory (DFT). A novel protein like-lectin Light Subunit Mushroom Tyrosinase (LSMT) was discovered inadvertently during elucidation of the button mushroom Agaricus bisporus tyrosinase structure. The molecular docking result revealed three possible positions, of which the first resembles the sugar-binding region in the structures of its homolog (HA-33 or CNL), and the second is located in the interface region to the tyrosinase subunit. Another position is a new finding region that includes interaction with five amino acid residues. The molecule complex was modeled by truncation of five selected residues, then the atom of peptide chain freezed. In the final study, the interaction energy was analyzed using DFT showed that Threonine 91 (Thr91) has the highest role of interaction between ligand and protein. Study at this fundamental level is important because it will be used as a benchmark of interaction characteristics between LSMT and Mannose. Thus, the calculation result can be a reference in the development of LSMT application as a drug carrier protein.

The objective of this study is to investigate the effects of experiential learning through farming on students’ learn- ing in the common core science GE course. A total of 101 year-one and year-two university students from different disciplines had joined the farming practicum, in which hands-on farm work and guided discussion were included. Quantitative surveys and qualitative feedback revealed that the farming practicum helped students understand the course materials better, foster their reflection on the environmental, social and political issues, and gain the knowledge and techniques in farming. This study provides insightful findings to support the implementation of experiential learning through farming in the university science GE course. Given the diverse aspects and interdisciplinary nature of agriculture, experiential learning through farming can be extended to other GE courses, including the humanity courses, to inspire the students and cultivate them into better global citizens.

This research aims to apply a series of classical machine learning algorithms based on decision trees (Decision Tree, Adaboosting, Bagging, Random Forest) to verify the ten-fold cross-validation of the steel plate fault data. The source of the data set was the Research Center of Sciences of Communication in Italy and has been used two times by M Buscema when it is provided [15, 16]. The data set includes 7 different types of steel plate faults: Pastry, Z_Scratch, K_Scatch, Stains, Dirtiness, Bumps, and Other Faults. It is found that the Bagging algorithm outperforms the other methods and achieves 96.30% and 90% accuracy on the training and testing set, respectively. This will allow us to find abnormalities on the surface of the steel plate timely and reduce losses. Based on these algorithms, we can cooperate with iron and steel practitioners to design more appropriate algorithms to achieve higher recognition accuracy in the future.

The article analyzes relaxation α-transition in elastomeric composites containing a hybrid carbon black/carbon nanotubes (CB/CNT) filler. According to the DMA data, the inclusion of hybrid particles CB/CNT in the filler leads to the expansion of the temperature dependences of the loss tangent (TanD) for all samples towards lower temperatures and the displacement of the position of the TanD maximum by a value from 4.0 up to 16 degrees in comparison with control vulcanizate. The DSC data indicate the presence of additional low-temperature α-relaxation transitions in modified vulcanizates (-123 ... -118 oC). The observed relaxation behavior of macromolecules is due to the appearance in the material of regions with less dense packing of macromolecules and, as a consequence, the expansion of their conformational set for segmental motion under low-temperature conditions. It guarantees to get material with increased fatigue resistance and frost resistance. It is shown that hybrid filler particles, in comparison with carbon black, change the final structure of vulcanizates, causing the appearance of areas with a more loose packing of macromolecules and increased segmental mobility in the low-temperature area. It gives opportunities for getting rubber with increased fatigue resistance and frost resistance.

The contribution focuses exclusively on volume manufacturers because the spread of sustainability effects is more limited for this market segment than premium brands. An empirical study is used to determine the expectations on the customer side regarding more sustainability in the automotive industry in general and in the interior sector in particular and to derive corresponding challenges and potentials for original equipment manufacturers. The empirical study is based on an online survey with randomly selected persons via social media. The survey was conducted via Survey monkey. All persons with a minimum age of 18 years were considered. A majority of 63 percent of all volume brand customers are willing to pay a higher price for additional costs related to the use of natural and sustainable materials. All participants principally agree that sustainability should play an overall role in the automobile and should not be restricted to individual areas. Further research projects will analyze whether there are significant differences when it comes to premium manufacturers. Furthermore, no distinction has been made between age, gender, and automotive brand. All of these can be explored in the scope of further research activities, too.

This study was done with the intent to screen the acetylcholinesterase activity of alkaloids from plant extracts often selected indigenous plants in Mindanao. Determination of the functional groups which may attribute to the anticholinesterase activity was done using the Fourier Transform Infrared Spectroscopy (FTIR). The presence of nitrogen, carbonyl group, hydroxyl groups, and aromatic rings was found in Costus speciosus, Euphorbia hirta, Ipomoea aquatica and Mimosa pudica. The results of the acute oral toxicity test show that at 2000mg/kg, Costus speciosus, Euphorbia hirta, Ipomoea aquatica and Mimosa pudica extracts are classified as Category 5: practically non-toxic based on the Globally Harmonised System [1]. Another purpose of this study is to acquire the Approximate Inhibitory Concentration (AIC50) of the extracts using the Ellman method to determine the% inhibition at concentrations logarithmically determined, and linear regression was used to calculate the AIC50. The AIC50 of C. speciosus, E. hirta, I. aquatica and M.pudica extracts are 4.18 mg/mL, 3.74 mg/mL, 3.68 mg/mL, 4.18 mg/mL, respectively and their corresponding AIC50 range are 3.13-5.23 mg/mL, 2.81-4.68 mg/mL, 2.76-4.60 mg/mL and 3.13-5.23 mg/mL. Furthermore, the AIC50 of the plant extracts was compared statistically with the positive control (Donepezil) to know if there is a significant difference. Results show that the significant value 0.287 is greater than α = .05 (2-tailed). The researchers failed to reject the null hypotheses that there is no significant AIC50 difference between of C. speciosus, E. hirta, I. aquatica and M. pudica extracts and the positive control, Donepezil. Hence,the anticholinesterase activity of the plant extracts are comparable with the anticholinesterase activity of positive control, Donepezil.

Lettuce (The current research was conducted to find the nitrogen content of fertilizer from coconut wastewater with microbial L. bulgaricus, L. acidophilus and S. thermophilus. The study was prepared based on a Completely Randomized Design with a single factor treatment. The treatments are P1: ZA 1.8 g/polybag Fertilizer, P2: Liquid Wastewater Fermented coconut water 250 ml/lt. water, P3: Fertilizer liquid wastewater of coconut water 500 ml/lt. water, P4: Fertilizer liquid wastewater of 250 ml fermented coconut/lt. water and R1M solution 10 ml/lt. water and P5: Liquid wastewater of fermented coconut water 500 ml/lt. of water and R1M solution 10 ml/lt. water. Each treatment was repeated as many as 3 replications. The growth observations included leaf area, wet weight, and dry weight of lettuce plant. The results showed that Inorganic fertilizer treatment and treatment of coconut wastewater fertilizer fertilized by L. bulgaricus, L. acidophilus, and S. thermophilus have no significant effect on leaf area formation, wet weight, and dry weight of lettuce plant. Coconut wastewater fertilizer is fermented by L. bulgaricus, L. acidophilus, and S. thermophilus concentration of 250 ml/lt. of water able to spur the growth of lettuce plant. An increase in waste output will accompany increased advancement in the coconut food industry's technologies. The waste of coconut water that is prevalent in the marketplaces emits an odor. Coconut water is an excellent media if it is used for the development of microbes.

The purpose of this rearch is to know the effect of the combination of biological agents with inorganic fertilizers and finding the correct dose for the combination of biological agents with inorganic fertilizer for the growth of sugarcane. The research was conducted in the village Pakiskembar, Pakis, subdistrict Malang. The study used a randomized block design with 7 treatments and 4 replications. Data were analyzed using Analysis of Variance (ANOVA); if there is a significant difference, then it is followed by a further test of LSD 5%. The results showed the effect of the combination of inorganic fertilizer with biological agents. There are significant differences in plant height and length of the rod at 139 DAP, stem diameter at 153 DAP, and the number of tillers at 97 DAP. P5 (NPK 300 kg ha−1 + ZA 400 kg ha−1 + Biofertilizer 30 L ha−1) is known to give good growth in most but not significantly different from P1 (NPK 400 kg ha−1 + ZA 600 kg ha−1). A combination of inorganic fertilizer with biological agents can lower the dose of inorganic fertilizer, and biological agents can provide nutrients needed for sugarcane so that the use of inorganic fertilizers can be reduced.

This study evaluates the performance of an imidazoline commercial inhibitor in a sweet environment. AISI 1045 carbon steel was chosen, with pH 5 and pH 7, flow rate 7.85 cm/s and 13.09 cm/s, and NaCl 3.5% solution. FTIR, XRD, Weight loss, Polarization, and Electrochemical Impedance Spectroscopy (EIS) tests were performed to obtain complete information regarding the inhibitor’s performance. According to weight loss results at pH 5, the highest efficiency of inhibitor was 82.59% with 200 ppm inhibitor’s concentration, flow rate 7.85 cm/s, and corrosion rate of 0.104 mm/y. While at pH 7, the highest efficiency obtained was 92.697% in 100 ppm concentration of inhibitor, flow rate 7.85 cm/s, and corrosion rate 0.037 mm/y. XRD testing showed Fe24N10 compound formed as a result of a reaction between Fe and the pyridine nitrogen atom. FTIR testing showed a functional group of inhibitors precipitated on the sample’s surface when immersed, and EIS testing showed that the addition of the inhibitor concentration increased the value of polarization resistance of solution, and the value of (constant phase element) decreased. Organic inhibitor becomes alternative protection of corrosion because it is biodegradable, cheap, and also environmentally friendly. The imidazoline inhibitor acts adsorptively on the surface of the AISI 1045 steel by forming the complex Fe24N10 to give a thin film that inhibits the rate of corrosion. Alternative corrosion inhibitors from tobacco can be used or even replace imidazoline inhibitors in applying oil and gas.

The main objective of the present study is to investigate the effect of natural coagulants (selected seed) on the treatment of domestic grey water and to study the factors affecting the removal efficiency in terms of Turbidity and COD removal. Effect of settling time and pH variation on the removal efficiency was also studied. Grey water is all wastewater generated in households, public or commercial properties without fecal contamination. Treating and reusing grey water decreases the quantity of fresh water needed and lessens the wastewater flow flowing into the sewer system. The grey water sample was taken from a residential society in Ahmedabad and the natural coagulant used was freely and locally available Custard Apple (Annona Reticulata) seeds. The initial characteristics of the grey water were accessed based on the parameters pH, TS, TDS, Turbidity, and COD and measured as 7.79, 780 mg/L, 590 mg/L, 276 NTU, and 625 mg/L, respectively. Turbidity and COD were analyzed for different Coagulant Dosage concentrations and by varying the Settling Time and pH. The settling time was varied from 10 to 30 minutes considering the elimination of micro-flocs, and pH was varied from 4 to 9. The optimum results were obtained at a settling time of 30 minutes, pH 9 and 10 mg/L coagulant dosage giving 85% turbidity removal and 82% COD removal. It was observed that adding organic coagulant beyond the optimum level contributes to COD. Water scarcity along with climate change, population growth, and development, pose difficulties for the present water supply systems. Today, 2.1 billion people globally are living without a safe water supply near their homes. Hence, domestic wastewater treatment along with its reuse is becoming a significant topic for research.

This study was carried out to assess the response of young cork oak plants inoculated by nine fungal isolates (Russul sp. Lactar controversus, Amanita pantherina, Cortinarius sp., Hebeloma sp., Boletus sp., Lactarius volemus, Inocybe sp. and Scleroderma sp.) to water deficit. Acorns were used as plant material. They were provided from canton "A" of the Maamora forest, harvested in Decem ber, and soaked in water for 24 hours before planting. As fungal material, 9 species of ectomycorrhizal fungi (Russula sp., Lactarius volemus, Lactarius controversus, Inocybe sp., Scleroderma sp., Amanita pantherina, Cortinarius sp., Hebe loma sp., Boletus sp) have been used for the prepa- ration of the inocula. The results showed that controlled mycorrhization significantly improves plants tolerance to drought stress. Boletus sp. was the most efficient isolate that procured to cork oak seedlings a better stomat- acal conductance, root and shoot dry weight, and chlorophyll content. Also, the leaf water potential, proline, and anthocyanin accumulation were lower in seedlings with Boletus inoculation. After the drought stress stage, cork oak plants have been rehydrated, and again, Boletus sp. produced a mean recovery of 60% while it was only 1% in plant control. These data clearly show that the inoculation of cork oak plants with ectomycorrhizal isolates, such as Boletus sp. could be a very interesting pathway in the sandy soils of Maamora and subsequently in determining the success of its regeneration programs.

This study aimed to investigate the effect of FUSO with various flow rate modulation patterns, amplitudes, and switching times on the conversion of synthesis gas. In this study, Forced Unsteady-State Operation (FUSO) was simulated by modulating the flow rate of the feed gas as a novel operation method to increase the conversion of synthesis gas to methane. FUSO was held after the system had reached its steady-state condition. The feed in this process was a mixture of H2, CO, and Ar with the mole fraction of 50%, 10%, and 40%, respectively. The average feed volumetric flow rate was 0.3 lN/min, and the operating conditions were 553 K with a total pressure of up to 2 barabs . Modeling and simulation for both steady-state and unsteady-state conditions were developed using the FlexPDE simulation software, assuming an isothermal fixed-bed reactor on Ni-based catalyst. The results were compared with bench-scale experimental data to validate the results. The simulation results compare sufficiently with the bench-scale experimental data to be validated. The simulation results showed that FUSO operation improves synthesis gas to methane conversion considerably. The highest conversion achieved was 84.79% under step pattern flow rate modulation with the amplitude of 0.6 lN/min and switching time of 3s, while the conventional steady-state operation only reached 72.99%. This study shows that FUSO can potentially be applied to synthesis gas methanation to increase conversion. Conversion of biomass into Synthetic Natural Gas (SNG) via gasification and methanation could be a favorable route to increase the potential use of biomass. The advantages of SNG are the high energy content and efficient end-use technologies, further supported by well-established gas distribution infrastructures.

A Trial Master File (TMF) plays a significant role in clinical trials to comply with regulatory requirements. Many organizations involve clinical trials in the process of implementing Enterprise Content Management (ECM) called electronic eTMF. The concept of this study was based on concerns about the implementation of the eTMF system. This case study aimed to identify and gain in-depth understanding of the gaps in eTMF system implementation for organizations in Thailand. Interviews followed a semistructured questionnaire with fourteen users of the eTMF system, which were conducted to collect users’ opinions concerning eTMF system satisfaction, tool (named master list) application, eTMF management processes complexity, and user knowledge. Besides, the survey was conducted to assess user knowledge on master list application and eTMF management processes. The interview findings revealed that unsatis- fied search function of the eTMF system was the highest ranking gap. The second and third ranking gaps were the lack of understanding and the complexity of the master list, respectively. Lack of knowledge concerning the master list and eTMF system functions was ranked as the fourth-ranking gap. Lastly, repeating the processes of eTMF was the fifth-ranking gap. Also, the results of the user knowledge assessment con- firmed the lack of knowledge in eTMF management processes and the master list. Determining the critical gaps in system implementation would be useful for the application of suitable solutions. Future study will utilize practices, tools, and guidelines developed for success in system implementation with a knowledge management approach. An efficient knowledge management program allows the utilization of knowledge, including improving users' productivity and satisfaction.

Plant Growth Promoting Bacteria (PGPB) have been considered as beneficial microorganisms that can promote plant growth by their potential mechanism, particularly Indole-3-Acetic Acid (IAA) production. Lysinibacillus fusiformis UD 270 isolated from local plants in Khon Kaen, Thailand displayed its capability to produce IAA from tryptophan well-known as its precursor. Thus, the study was investigated in IAA biosynthesis with different concentrations of tryptophan and bacterial cell number as well as observed the correlation between bacterial growth and IAA production. The bacteria were grown in nutrient broth supplemented with 0, 0.5, 1, 2, 3, 4, and 5 mg mL-1 of tryptophan at 30oC for 72h. Reasonable bacterial growth and IAA production were appreciably determined at 36h of incubation in 5 mg mL-1 of tryptophan. Meanwhile, the pH of the culture media was gradually increased at the beginning and made stable after 36h. Moreover, when the cell concentration was varied from 103 to 109 Colony-Forming Unit (CFU) mL-1, the IAA production was significantly higher at high cell concentrations, concretely with 107 and 109 CFU mL-1. Also, a positive correlation between growth and IAA production was indicated in the two experiments. The results have referred for the suitable uses of tryptophan and cell number for bacterial growth and IAA biosynthesis of L. fusiformis UD 270 to apply for further researches.

Mercury is a toxic pollutant emitted from industrial sectors to the environment and distributed globally. The potential for biological treatment of industrial wastewater contaminated with mercury was evaluated using Pseudomonas putida (P. putida) under various conditions in a bioreactor. The effect of mercury concentration on the P. putida growth of bacteria and also mercury removal was determined. Modifications in optimum operating conditions in shake flask and bioreactor need to be determined so it could bring us to a better result. In this research, optimum conditions for growth of P. putida in shake flask are identified: acclimatization time 24 hours, orbital shaker speed 180rpm, temperature 37째C, pH 7, and nutrient concentration 8g/L. The removal efficiency obtained is 99% for 1ppb, 99.8% for 6ppb, and 98.6% for 19ppb while for 1000ppb mercury, the removal efficiency is 92% for 1 hour and 98% for 28 hours. In 2L bioreactor, same condition as shake flask is applied with agitator speed of 180 rpm and aeration time of 0.50vvm. For 1300ppb and 3000ppb, the removal efficiency is 89% and 94%, respectively. The findings of this study can be used as a reference for future application in industrial wastewater treatment plant.

This research purpose was to identify attention as the primary characteristic of mindfulness, among other cognitive features. The utility of training attention is evident in real-life situations such as listening to others, driving a car, conducting a medical-surgical procedure, and so forth. Therefore, we argue that devising a method for detecting the moment at which the attention is distracted would be beneficial to the cultivation of attention. We have conducted research to develop a software framework that can model attention pertaining to a particular task and give an alert when attention is distracted. The framework has been designed to capture attention-related Electroencephalography (EEG) brain wave signals in response to a specific task and to train an Artificial Neural Network (ANN). The trained ANN can be used to receive EEG signals during a task and determine an individual's attentiveness. Accordingly, a vibration alert is sent to an individual's mobile phone to serve as a signal for the person to refocus attention. The framework has been used to model attention during a lecture, and an experiment was conducted to assess the attentiveness of students. The experimental results determined that 75% of students were able to maintain attention during a lecture, and vibration alert has been effectively supportive of regaining the attention. Hence, we conclude that our software framework can be used to the model regaining attention in a session that requires the focused attention.

This study applied the Markov chain model on the daily average wind speed data recorded at the meteorological stations in northern Peninsular Malaysia. This study aims to investigate the trend of wind speed by obtaining the transition probability matrix and the stationary distribution vector for each of the stations. The five states of wind speed based on the Beaufort scale ranging from the scale Beaufort 0 up to Beaufort 4 were defined. The stationary distribution vectors obtained revealed that Kota Bharu, Kuala Terengganu and Bayan Lepas demonstrated the highest proportion of daily average wind speed occurring in the scale of Beaufort 2 with the proportion of 69.27%, 63.62% and 61.89% respectively. Meanwhile, Alor Setar and Chuping showed the highest proportion of daily average wind speed occurring in the scale of Beaufort 1 with the proportion of 54.52% and 72.29% respectively. Furthermore, Kota Bharu and Kuala Terengganu also showed 9.30% and 7.13% proportion of daily average wind speed occurring more than 3.3 meter per second (Beaufort 3 and above) while Bayan Lepas station only demonstrates approximately 3.31% of the category. The least proportion displayed for this category is Alor Setar with 0.6% and followed by Chuping with 1.98%.

This study aims to calculate the angular velocity, angular acceleration, and torque of two rigid bodies that share a common point using instantaneous sensor measurements. The goal of this approach is to produce outcome measurements free from accumulated error. IMU sensors and accelerometers have been used to capture real-time data. This arrangement applies to cases such as when the bodies are connected by a ball-and-socket joint, a Hooke joint, or a revolute joint, especially where it is impractical to use a joint measurement sensor between the bodies the relative motion of human limbs. The proposed system has been designed, built, and tested in a lab, which showed satisfactory results. The proposed sensor assembly overcomes these limitations. The case study showed accurate data capturing without any limitations. This paper has also demonstrated various formulas and algorithms that can help to calculate both robot parameters with these different sensors, which help to find out angular acceleration, torque, and angular velocity for smooth and reliable opeerations. The proposed circuitry system and algorithm showed very accurate results of these parameters.

The objective of this study was to investigate the interaction of different types of cement and superplasticizer admixtures in which the Superplasticizers were stimulated by heating before addition to the mortar. To evaluate the desired impacts, tests on the fluidity and flow loss of mortar were conducted. For better considering the behavior of admixtures under heat stimulation with different types of cement, three types of cement and two types of superplasticizer are used, Polycarboxylic acid-based and Polycarboxylic acid-based ether. The admixtures were heated to 40 ̊C, 50 ̊C, and 60 ̊C for 0.5 hours and 24 hours. Generally, Polycarboxylic acid-based admixtures showed higher flow improvement in 60 ̊C heating with all cement types, especially with High Early Cement, in comparison with Polycarboxylic acid-based ether type. The flow loss of mortar was reduced by heat stimulation of admixtures, Polycarboxylic acid-based ether (the product suitable for Ready Mix Concrete) showed significant flow improvement after 15 minutes and kept it to one hour. As the result of this research, it is possible to reduce the admixture dosage and maintain the desired fluidity of mortar or increase the mortar fluidity by applying the heat stimulation technique.

Therefore, this study aimed to isolate and screen PNSB from shrimp ponds with their ability to reduce phosphate in water from shrimp cultivation. A total of 83 PNSB strains were isolated from water and sediment samples collected from various 15 shrimp ponds located in Phang-nga and Songkhla provinces. For primary screening, there were 42 strains (51%) that grew well (OD660> 1.0) in glutamate-acetate broth supplemented with 1.5% (w/v) NaCl, under conditions of microaerobic-light and aerobic-dark. However, in secondary screening, only two strains (W12 and W48) could grow in sterile rearing water collected from shrimp ponds. They were selected for tertiary screening to investigate their ability to remove phosphate in sterile rearing water under both incubating conditions. Both PNSB strains produced no significant differences for phosphate removal efficiency (> 50%) with the exception under microaerobic-light conditions as strain W12 roughly reduced 46% phosphate. Of these, 2 strains could be used as inoculants to remove phosphate from rearing water in shrimp ponds. One of the key environmental concerns about shrimp cultivation is the discharge of rearing water with high levels of nutrients, especially phosphate, into waterways, resulting in eutrophication. To solve this problem, biological treatment is well-recognized, and the use of purple non-sulfur bacteria (PNSB) is one of the attractive alternative choices because of their high removal efficiency in wastewater treatment with various metabolic growth conditions.

In this paper, we design and implement a two-axis tracking system of PV systems that follows the Maximum Power Point (MPP) using a programmable circuit XILINX type Complex Programmable Logic Device- CPLD and Xilinx ISE software. Thus, the PV module will reach its MPP in relation to the date and time of the day. The test bed relies on an algorithm integrated in the XILINX that has as inputs: date, location’s latitude and longitude, the standard longitude (related to the location’s position in relation with Greenwich), and the number of the positions of the Sun's path. To establish the position of the panel in a time of day value is determined by the following calculations: the angle of the day, a correction factor of Earth's orbit, the solar declination angle, the equation of time in minutes, eastern time using latitude angle, the number of hours the Sun shines using angle eastern time, time the Sun sets, vectors containing the coordinates of the positions of the Sun (in this case 10 positions) during the day and azimuth angle. A Photovoltaic (PV) tracker system is one of those methods that are able to increase PV power generation. Theoretically, a PV tracker system with two axes can increase the overall solar energy capture by about 45%, compared to a fixed PV module tilted at an angle equal to the local latitude. For a one-axis tracking system, the increase is approximately 32%.

The paper aims to explore how South African retail organisations can use m-commerce to achieve their strategic objectives. Strategic objectives include increased return on capital, improved operating quality and efficiency, growing high-value customer relationships, accelerating product innovation, and creating a high-performance culture. Research questions were generated from a strategy map that identified five strategic themes, each with its own objectives. A qualitative approach to the research was taken with open-ended interview questions, and thematic analysis was used to analyse the data. The literature and the data analysis findings indicated that South African retail organisations could use m-commerce to achieve all the strategic objectives. However, the findings indicate that m-commerce, within the South African retail context, is more suited to increasing return on capital, growing high-value customer relationships, and improving operating quality and efficiency. The literature surveyed highlighted that there could be an opportunity for retail organisations within South Africa to achieve their strategic objectives by implementing effective m-commerce strategies which are aligned with the business strategy. Therefore, improving and updating their business processes and modifying

The search for p-type Transparent Conducting Oxides (TCO) has been pursued vigorously to complement current advances in n-type TCOs to develop heterojunctions for various electronic devices. One of the promising p-type TCOs is CuAlO2. Thin films of CuAlO2 were deposited on a clean glass substrate using the chemical solution deposition (sol-gel) method of deposition with CuCl and AlCl3 taken as the starting materials. CuCl was dissolved in HCl while AlCl3 in distilled water, pH value of the mixture was controlled by the addition of NaOH. The samples were annealed at different temperatures in order to determine the effect of annealing temperatures on the morphological and optical properties of the deposited CuAlO2 thin film. The surface morphology reveals an improved crystalline nature as annealing temperature increases. The UV-vis and FT-IR spectrophotometry results indicate that the absorbance for all the samples decreases sharply from a common value of about 89% at about 329nm to a range of values of 56.2%-35.2%. The extinction coefficients for all the samples decrease from 133.89 x 10-3, 111.76 x 10 -3, 93.45 x 10-3 and 89.44 x 10-3 in the infrared region to about 81.11 x 10-3, 82.22 x 10-3, 83.35 x 10-3 and 84.42 x 10-3 at about 4.05eV in the visible region. And the absorption coefficients of three samples decrease with increase in annealing temperature from 1.58 x 10-6, 1.29 x 10-6 and 1.08 x 10-6 at about 1.14eV in the infrared region to about 1.93 x 10-6, 1.58 x 10-6 and 1.29 x 10-6 at about 3.62eV in the visible region. Transmittance and band gaps vary directly with annealing temperature; the deposited films were found to be suitable in optoelectronic applications.

The main aim of this research is to study the effect of silver as substitution to super-conductivity behav ior of Hg_0.6Tl_0.4Ba₂Ca₂(Cu1-xAgx)₃O_8+δ compound through modifying phase and lattice parameters, transition temper ature Tc, likewise, the dielectric properties. Bulk polycrystalline Hg_0.6Tl_0.4Ba₂Ca₂(Cu1-xAgx)₃O_8+δ compound compound samples with x = 0.0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6 and 0.7, are synthesized by solid state reaction process. Study identifies Silver substitution on superconductivity behavior. Structural specifications are deliberated by using X-ray powder pattern, the high temperature phase superconductor (Hg-1223) of the tetragonal structure didn't change with the partial replacement of Cu+2 by Ag+1 ions, lattice parameters a, c ,c/a are established to vary as function of Ag- substitution. Transition temperature (Tc) has been calculated using technique of four-probe to measure electrical resistivity. Transition temperature at zero resistivity T c(of fset) decreases from 117 to 86 K with increasing Ag. In addition, dielectric properties (dielectric constant and loss) are characterized directly by relating with Ag concentration.

This research will present the procedure to obtain the storage capacity, the type of LNG receiving terminal, and Critical Success Factors (CSF) as well as control measures for such factors, which are appropriate for Phra Nakhon Tai power plant. This research methodology starts from determining the terminal’s storage capacity from the demand of natural gas of Phra Nakhon Tai power plant. Then, the storage capacity and the calculated natural gas demand will be used to select the type of LNG receiving terminal which is appropriate for the power plant. The selection criteria will come from a study of a group of countries with LNG receiving terminals and have similar sea state characteristics to Thailand and from reviewing related literature. Subsequently, when the type and the terminal's storage capacity are known, the next process will be to create measures by starting from analyzing the CSF for the establishment of the terminal by applying the concept of Balanced Scorecard and using those CSF to create control measures. Once measures of each factor are known, they will be evaluated to find the compatibility between the factor and the measures. The final results achieved will be the result of categorizing control measures of critical success factor into 3 phases for the establishment of a LNG Receiving Terminal in the Gulf of Thailand to supply natural gas to Phra Nakhon Tai power plant continuously. It was found that they can be categorized into 4 aspects, namely Metocean, logistics, environment, and economics. Hence, this research is the first research that will study all four aspects and analyze factors for LNG receiving terminal establishment in each aspect as well as their control measures.

Life is full of frustration and challenge that could lead to mental stress. Teaching could be one of challenging profession which has become crucial to the society. Teachers uphold such abundance work including teaching, handling students and activities at school. Despite of their overload work, the new developed technology need to be taught by the teacher. The technology is actually consists of critical thinking and hands on activity which requires the teacher to be creative in conveying the knowledge. Thus, an experiment has been conducted to identify the teacher’s stress by using electroencephalogram (EEG) signals and Depression Anxiety Stress Scale (DASS21). Affective Space Model (ASM) developed by Russell has been used in this research study which consists of valence and arousal. The study in understanding teacher stress by using EEG and ASM has been found as scarce. The objective of this research study is to identify the stress level of the teacher through their emotion. There are two experiments need to be conducted: first experiment is to profile the subject’s basic emotions and second experiment will be answering the Depression Anxiety Stress Scale (DASS21) to induce stress. 10 healthy teachers are recruited for this study. Mel Frequency Cepstral Coefficients (MFCC) is adopted for feature extraction and Multi-Layer Perceptron (MLP) will be used as classifier. The result shows that the emotions appear is more towards negative emotions which depicts stress for the subject after teaching the developed technology. Based on this results, we may do further research on the educational system in Malaysia if it is been embedded with the new technology. Furthermore, the study is beneficial for the early stress detection among teacher.

International Direct Dialing (IDD) is one of the services based on the Telecommunications Operator clear channel access and Voice over IP (VoIP). In running this business, Operators face Grey Operators who do illegal practices by passing traffic of international incoming call without going through the official international service providers called Fraud Subscriber Identity Module Box (SIMBOX). The impacts of this practice are not only the revenue decline, but SIMBOX also provides less good image for the operator because of the low quality service. Some operators have made efforts to implement the mitigation of traffic SIMBOX fraud detection system. This study aims to improve the detection of fraud traffic and maintain the quality of service. This study redesigns the existing SIMBOX fraud detection system to become a dynamic detection system by adding a dynamic control algorithm and is simulated using MATLAB simulation approach. A dynamic system is indispensable as there are various fraud traffic flow profiles that always change and could not be predicted. The results of this study indicate that fraud detection SIMBOX could be improved up to 5,000% and could increase potential revenue to $ 2 billion per month. Thus the fraud detection SIMBOX dynamic system will provide greater detection results than the previous system.

Bandwidth starvation is an important aspect to consider when deploying Triple Play Quality of Service (QoS) in Internet Protocol (IP) network. How to guarantee each class of QoS running smoothly with enough bandwidth when facing congested network. Traffic policing technique has been proposed by author to ensure voice traffic is separated from other class instead of using Low Latency Queuing (LLQ) of Weighted Fair Queuing (WFQ). This paper addresses the effect of traffic policing and Random Early Detection (RED) to quality of service in term of delay, jitter and packet loss. The result show traffic policing with combination using RED is promised each class of WFQ gets enough bandwidth and avoid bandwidth starvation. Simulation results show end to end delay for voice 0.02 second, voice jitter 0.00025 second,video delay 0.05 second,video jitter 0.0005. These paramaters accepted ITU-Y1541 standard. For queuingdelay paramater the result is 0.02 second.

Nowadays, along with the dramatic development of industrial automatic, optimization problem has been playing an important role in designing controllers for nonlinear systems. This paper proposes a hybrid control design of Fast Output Sampling Discrete Sliding Mode Control (FOSDSMC) and fractional order PID controller (FOPID) based on fireworks algorithm (FWA) to optimize controller parameters. The hybrid controller is verified on a nonlinear inverted pendulum system. The simulation of controller optimization process is carried out using MATLAB/Simulink. The results are compared with two published controllers such as a hybrid control design of PID controller and fast output sampling discrete sliding mode control, and a hybrid control configuration of PID and state feedback controller based on linear quadratic regulator method. The comparison results show the better performance of the proposed method.

Today manufacturing systems are highly automated and consist of several interlinked machines. These automated lines are subject to frequent failures, which affect system reliability and availability, as well as its productivity. In operation of such lines, it is necessary to have enough information on failure and repair data in order to be able to analyze system reliability and availability so that exact output rates could be estimated. Reliability also depends on the preventive maintenance operations. Therefore, it is also desirable to have appropriate analysis in order to see the effects of preventive maintenances on system availability. This paper presents a procedure for collecting appropriate data, analyzing it, and determining system reliability, availability, and productivity of manufacturing lines. Furthermore, procedures and models are presented to study the effects of preventive maintenances on system availability. A special case example is used to illustrate the analysis in detail. The procedures and the models presented in this paper should be useful for operations engineers in order to improve the productivity of their manufacturing lines.

The ASEAN region is affected by many climate-change inducing factors such as Forest Fires, Transboundary Haze Pollution, and Forest Degradation. However, within the next few years there may be one more significant contributor in making the ASEAN region vulnerable to climate-change–Air Transportation. Air Traffic Demand in the ASEAN region is growing tremendously. It is expected that air traffic in the region will triple by the year 2033, thus, posing the need for reduction of Green House Gas emissions from aircraft to reduce air pollution. This paper highlights the estimated amount of key air pollutants such as NO_X and CO₂ emitted in the ASEAN region due to current air traffic demand and the potential benefits of free route airspace implementation in the region. The environmental emissions were estimated using System for Traffic Assignment and Analysis at a Macroscopic Level (SAAM) tool. A simplified version of EUROCONTROL’s Advanced Emission Model was adopted for estimating the environmental emissions. This paper also discusses future ATM technologies that may be implemented in the ASEAN region, which could support the feasibility of the Free Route Airspace Concept in the region.

Internet of Things (IoT) is used for devices to interact with each other, and Femtocells are used to provide reliable communication by eliminating shaded areas where wireless signals have become weak. IoT security is crucial since the untethered nature of wireless networks primarily allows for eavesdropping threats to confidential information. Therefore, the interlock protocol is proposed to protect confidential information that is prone to eavesdropping due to the use of an unsecure public key. This paper addresses this limitation through a countermeasure that combines the time synchronization one-time password (OTP) and the interlock protocol. In the proposed method, we use OTP for authentication before transmitting the public key and data. In order to counter eavesdropping attacks, the OTP should be first used to detect the attacker. Simulations show that both methods have up to 46% of detection rate. However, our method has a prevention rate that is 54% higher than that of the interlock protocol.

Mean-Variance Mapping Optimization (MVMO) is the newest class of the modern meta-heuristic algorithms. The original version of this algorithm is suitable for continuous search problems, so can’t apply it directly to discrete search problems. In this paper, the binary version of the MVMO (BMVMO) algorithm proposed. The proposed Binary Mean-Variance Mapping Optimization algorithm compare with well-known binary meta-heuristic optimization algorithms such, Binary genetic Algorithm, Binary Particles Swarm Optimization, and Binary Bat Algorithm over fifteen benchmark functions conducted to draw a conclusion. The numeric experiments result proves that BMVMO is better performance

In this paper, the meta-heuristic algorithm which named Differential Evaluation (DE) has been improved. The improving made to increase the exploration rate and decrease the run time. Since DE needs too long time, when we implement it to solve computational expensive problems, we developed two different versions of DE named by Enhanced1 Differential Evaluation (E1DE) and Enhanced2 Differential Evaluation (E2DE). E1DE and E2DE were introduced to solve Computationally Expensive Optimization (CEO). Problems discussed and tested using all 15 test functions of the Special Session & Competition on Real-Parameter Single Objective Optimization (Expensive Case) at Congress on Evolutionary Computation 2015 (CEC-2015). The results show that the work significantly improved the basic DE in time by 54% and in results by 86%.

At COP 21, participants of the UNFCC reached an agreement for mitigation of greenhouse gas emissions. It is not mandatory, but new policies and technologies are needed for each country to accomplish Intended Nationally Determined Contributions (INDC). Especially in the power section, Smart Grid, Renewable energy, Battery Energy Storage System (BESS), Distributed Generation and Microgrid are emerged as solutions to reduce GHGs. These technologies are known as the GHGs mitigation Technologies. However, researches are needed to reveal that is true or not. BESS can be operated for diverse purpose. This paper presents the GHGs emissions changes resulting from the unit commitment with BESS that is applied to Korea power system for minimizing end-user’s costs. This study is based on IEEE 39-bus system to reveal the influence of BESS on GHGs. As a result, in a certain condition, the increasing of BESS capacity could result the increasing of GHGs emissions.

There are many spatial interpolations schemes, but none of them can perform best in all cases. Hence, this study aims to find an optimal interpolation scheme for precipitation in Selangor and Langat basin of which are the two major basins in Selangor. In order to obtain spatially distributed precipitation data, 21 measured rain gauges points are interpolated. Five interpolation methods have been tested after exploring data and cross-validation was used as the criterion to evaluate the accuracy of the various methods. The best method was obtained by the kriging method while the inverse distance weighting (IDW) perform worst.

In the present study, conventional staining and NORs banding as well as Fluorescence In Situ Hybridization (FISH) using the 18S rDNA and telomeric (TTAGGG)n probes were applied to stain the chromosomes of crocodile catfish, Bagariussuchus (Siluriformes, Sisoridae) from the Chao Phraya River, Thailand. Kidney cells of six male and six female crocodile catfishes were used as a sample. The mitotic chromosome preparations were done directly from kidney cells. The results showed that the diploid chromosome number of B. suchus was 2n=56, the Fundamental Numbers (NF) were 102 in both male and female. The karyotype comprises 17m+17sm+12a+10t. The Nucleolar Organizer Regions (NORs) were detected by Ag-NORs banding and 18S rDNA probe mapping. The 18S rDNA are terminally located on the short arm adjacent to the telomere of the single pair of the 1st chromosome pair whereas NOR-bearing chromosome is only one chromosome of the 1st chromosome pair (1a 1b, polymorphic characteristic) at the subtelomeric region of the short arm. Moreover, FISH with telomeric probe showed hybridization signals on each telomere of all chromosomes and interstitial telomeric sites were not detected. There were variations in signals of FISH and their position in the karyotype along with variation in DNA sequences. These markers are useful for future discrimination of population of closely related species and their polymorphism.

Eco-brick is an important environment-friendly strategy to mitigate carbondioxide emission in building construction. For energy and resource efficient of eco-brick manufacturing, natural fibre and biomass waste can be utilised as matrix. The present study aims to evaluate the properties of concrete eco-brick manufactured using kenaf fibre matrix. The brick was produced by mixing kenaf fibre (MR grade), Portland cement and sand. The physical and mechanical properties of the eco-brick were evaluated according to the ASTM C73 method. The water absorption values of the eco-brick were in the range of 9.0-12.5 %. Eco-brick produced from 0.5, 1.0, 1.5% and 2.0% kenaf fibre have flexure resistance to compression of 4693.9, 4335.6, 3879.3 and 3294.5 psi respectively. According to the ASTM C73 guidelines, the properties of eco-brick from mixture of 0.5- 1.5% kenaf fibre produced in this study meet building material criteria for construction in moderate weather conditions.

The modified approach on fuzzy time series, which is represented by [3], is applied to rubber price in Thailand. The developed forecasting method corresponds to the uncertain data. The nearest symmetric trapezoidal fuzzy numbers are used to further enhance the forecasting accuracy. The accuracy of this method is compared to the traditional method and actual values by the Mean Absolute Percentage Error (MAPE). The results show that the forecasts by the developed fuzzy time series forecasting method is more accuracy than the traditional method.

Northern Kazakhstan is the basic region producing grain in the Republic of Kazakhstan. National economy depends on its state. A brief description of the characteristics of soil and the climatic conditions is given in the article. The results of years’ research (n=600) in diagnostics and optimization of crops’ nitrogen nutrition in the conditions of insufficient and unstable moistening of Northern Kazakhstan are considered in this article. The methodological approach to the evaluation of soil nitrogen state and crops with nitrogen is explicated. New method of determining the needs and dose calculation in nitrogen fertilizers with individual requirements of crop and the main factors defining their effectiveness is represented. The method is based on identification of major factors of soil fertility defining the efficiency of crops and nitric fertilizers. On the basis of the correlation analysis, the quantitative interrelation of factors with efficiency of crops is defined. Optimum parameters for measuring the content of nitrogen in the soil required for the maximum efficiency of crops and ways for achieving it are determined. The method considers individual requirements of crops and the main factors defining effectiveness. The developed technique allows purposefully managing soil nitrogen regime supplying optimization of nutrition and implementation of crops’ potential possibilities.

Celosia plumosais an attractive ornamental plant having flame-like flowers and is used traditionally as a medicinal herb. This paper deals with thered-coloured callus induction from the root, stem and leaf explants when cultured on Murashige and Skoog (MS) medium supplemented with different concentrations of auxin picloram and 2,4- dichlorophenoxyacetic acid (2,4-D), applied singly. When explants were cultured on MS media supplemented with 6.0 mg/l picloram, the highest amount of red-coloured callus (0.49±0.26 g) was from leaf explants and the lowest (0.09±0.02 g) from stem explants. However, the highest amount of red-coloured callus (0.69±0.13 g) was achieved on Murashige and Skoog (MS) media supplemented with 1.5 mg/l 2,4-D from stem explants, while the lowest (0.13±0.06 g) from root explants was cultured on 0.5 mg/l 2,4-D. Generally, the optimum concentration for red-coloured callus formation using picloram (6.0 mg/l) was higher compared to 2,4-D (1.5 mg/l 2,4-D).

Epiphytic terrestrial green algae are normally favoured by an environment with higher pH level. Air pollution in the atmosphere contributes to altering the bark pH and provides a better medium for algal growth. High absorption capacity of the microalgae makes it easy to accumulate atmospheric pollutants in their cells immediately. Habitats of epiphytic terrestrial algae are mostly characterized by aridity, and/or levels of temperature and light intensity. Bark pH of tree surfaces has been considered as one of the most important factors affecting the community structure of corticolousbiocells. The present work was designed to assess the effect of bark pH on the number of algal cells inhabiting 15 standing trees from the sampling station located in the Central Region of Peninsular Malaysia, Putrajaya. Several methods were used including field sampling, algal quantification, algal identification and measurement of bark pH. The study revealed that the density of epiphytic terrestrial algae was found to be significant with the bark pH (p- value= 0.001). This positive correlation (r-value= 0.762) showed that bark pH does play an important role in the health of algal cells. The algae are believed to be able to tolerate higher bark pH. The alkaline bark pH altered the microalgal composition because it was found to be positively affecting the density of epiphytic microalgae. Therefore, higher bark pH significantly contributes to the enrichment of algal density

This study documents the traditional fishing gears and fishing methods of the rural Ilocano fishermen of which this has not been done comprehensively. It records, identifies, and describes their characteristics that include their designs, mode of operation, fishing grounds and species of fish caught, among others. Artisanal fishermen using the fishing gears and fishing methods from the inland and coastal municipalities of the province were the sources of information. Results of the study revealed that there are 48 fishing gears and fishing methods used which are classified into four (4) categories based on the classification of fishing gears in the Philippines. There are eight of the hand instruments (six coastal, two inland and three common); ten of the traps (three coastal, seven inland and one common); 14 of the lines (11 coastal, two inland and one common); and 16 of the nets (10 coastal, six inland and four common). Fifteen miscellaneous fishing accessories or paraphernalia which are not in the classification were also documented. A variety of marine and freshwater fish species comprising the catch of the fishing gears and fishing methods wereidentified and recorded in their Ilokano and English or common names. These include 44 species of marine fishes, six (6) species of marine invertebrates, seven (7) species of freshwater fishes, and three (3) species of freshwater invertebrates. A documentation of these traditional implements is valuable material to preserve the fishing culture and traditions of the Ilocanos. Researchers and other interested persons may find this study as an important material for further studies.

With the growing demand for biofuels, ethanol production is rising. Alternative energy crops have been investigated to get better yield from little resources. Sweet sorghum and sweet pearl millet are promising energy crops. However, the sugar is mainly located in the juice rather than in the grain. Usually, the biomass of these crops is carried to a plant where it is handled like that of sugarcane. With the rise of the transportation fees, carrying the biomass leaves less profit to the producer and causes the loss of organic matter or forage. The objective of the research study was to design, build, and test an in-field mobile juice extraction prototype press. This allows pressing on-the-run the biomass harvested with a forage harvester. The pressed material (bagasse) is dumped on the ground while the juice is collected. The prototype press was built in the summer of 2014 and preliminary tests were carried out in the field. Obtained results are promising as 57% of the total water was extracted. More tests will be carried out to optimize the prototype press.