An accurate energy consumption prediction becomes crucial with increasing electric vehicle usage for effective power grid management. This research examined the performance of eleven machine learning models for this purpose: Ridge Regression, Lasso Regression, K-Nearest Neighbors, Gradient Boosting, Support Vector Regression, Multi-Layer Perceptron, XGBoost, CatBoost, LightGBM, Gaussian Processes for Regression(GPR) and Extra Trees Regressor, considering real historical data from Colorado. The models were evaluated using different metrics: Mean Absolute Error (MAE), Mean Squared Error (MSE), R², Root Mean Squared Error(RMSE) and Normalized Root Mean Squared Error(NRMSE), with visual analyses through scatter plots and time series plots. The best model observed was the Extra Trees Regressor, which had an MAE of 0.5888, an MSE of 3.2683, R² value of 0.9592, RMSE of 1.8078 and NRMSE of 0.020. Gradient Boosting and KNN also returned good results, although they were slightly more dispersed. Nevertheless, while non-linear models like MLP, XGBoost, CatBoost, LightGBM and linear models such as Ridge and Lasso Regression offer valuable insights, they exhibit shortcomings in estimating energy, especially at extreme levels, highlighting limitations in capturing complex non-linear interactions. This study focuses on their applicability to energy projections to demonstrate how well ensemble and non-linear models may capture intricate patterns in time series. These cutting-edge machine learning techniques might greatly enhance energy demand predictions.

This study presents a wind speed prediction model using monthly average wind speed data, employing the Extreme Gradient Boosting (XGBoost) algorithm to enhance forecasting accuracy for wind farm operations. Accurate wind speed forecasting is crucial for optimizing energy production, ensuring grid stability, and improving operational planning. Existing studies on enhancing wind speed prediction using ML algorithms have some drawbacks based on accuracy, efficient prediction, and stuck-in-local-optima parameters. The dataset comprises monthly average wind speed measurements, and extensive preprocessing is conducted to prepare the data for machine learning. Various hyperparameter tuning techniques, including Randomized Search, Grid Search, and Bayesian Optimization, are applied to improve prediction accuracy. The performance of the model is evaluated utilizing key metrics, including Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), Mean Absolute Percentage Error (MAPE), Continuous Ranked Probability Score (CRPS), and Maximum Error. The results indicate that hyperparameter tuning significantly improves model accuracy. Specifically, Grid Search demonstrates superior performance for short-term (one-month) forecasting, while Randomized Search is more effective for long-term (six-month) forecasting. The findings emphasize the critical importance of hyperparameter tuning strategies in the development of reliable wind speed forecasting models, which have significant implications for the efficient management of wind energy resources.

The global transition toward electric vehicles (EVs) necessitates precise forecasting of charging demand to ensure grid stability and optimize infrastructure. Although gradient boosting (GB) Machines demonstrate significant potential for this task, performance limitations often arise from inadequate hyperparameter tuning. However, existing studies lack a systematic comparison of advanced metaheuristic hyperparameter optimization techniques and stacked ensemble learning under a unified forecasting framework for EV charging demand prediction. This study introduces a meta‐optimized forecasting framework that integrates advanced hyperparameter optimization techniques with a novel stacked ensemble architecture. A comprehensive comparison of five optimization methods, covariance matrix adaptation evolution strategy (CMA‐ES), Bayesian optimization (Hyperopt), particle swarm optimization (PSO), Optuna, and Randomized Search, is conducted to tune a GB model. Analysis indicates that CMA‐ES yields superior performance among single models, achieving a reduction in mean absolute error (MAE) by 23% compared to default parameters. Furthermore, the proposed stacked ensemble model amalgamates predictions from diversely tuned base learners through a meta‐learner, resulting in a breakthrough MAE of 0.94 kWh (a 53.6% reduction) and an accuracy measure of R ² value of 0.95. This ensemble effectively eliminates extreme errors and exhibits exceptional temporal fidelity, demonstrating responsiveness to demand spikes with minimal lag. The findings offer clear guidelines for selecting optimization strategies based on accuracy–computational cost trade‐offs, establishing that systematic hyperparameter optimization in conjunction with ensemble learning represents a transformative approach for EV charging demand forecasting.

This study presents a robust, end-to-end machine learning framework for detecting Non-Technical Losses (NTLs) in smart power grids. The proposed Optimized Bootstrapped Gradient Classifier (OBGC) integrates k-Nearest Neighbor (kNN) imputation for missing data handling, Adaptive Synthetic Sampling (ADASYN) for class imbalance correction and Optuna-based hyperparameter optimization. The framework is evaluated using the State Grid Corporation of China (SGCC) dataset comprising 42,373 consumers over 1035 days (January 2014–October 2016), characterized by 25.6% missing values and a severe class imbalance (91.46% honest vs. 8.54% theft-prone consumers). To ensure robust evaluation, all experiments employ stratified 5-fold cross-validation on the training set, with a held-out 20% test set reserved for final performance assessment. Experimental results show that OBGC achieves an accuracy of 96.61%, precision of 96.18%, recall of 97.13%, and F1-score of 96.65%, substantially outperforming baseline XGBoost (accuracy 92.75%, F1 36.11%) and other state-of-the-art classifiers. An ablation study confirms that ADASYN is the most critical component; its removal causes recall to collapse from 97.13% to 16.87%, demonstrating the severe impact of class imbalance on theft detection. Optuna-based tuning contributes an additional 3.29 percentage-point gain in accuracy over the unoptimized baseline. The framework offers a transparent, reproducible, and operationally feasible solution for real-world NTL detection.

Hydrogen (H ₂ ) has attracted considerable attention as a promising energy carrier due to its high energy content compared to conventional fossil fuels. The continuous growth of the global population is driving an increasing demand for energy, leading to higher consumption of fossil fuels and, consequently, increased greenhouse gas emissions and climate change. This has driven the research community toward renewable energy sources (RESs) such as solar, wind, and tidal energy. Among RES options, hydrogen stands out as one of the most viable alternatives. The significant research has focused on hydrogen production and its applications, comparatively less attention has been given to hydrogen storage. In this context, the hydrogen storage characteristics of alkali metal (AM) and alkaline earth metal (AEM) decorated and defective hexagonal boron nitride (h‐BN) nanosheets were systematically investigated using density functional theory (DFT) calculations. The AM atom (Li) and AEM atom (Be) are found to be uniformly dispersed on h‐BN surfaces containing boron vacancies and divacancy defects, effectively preventing atom clustering. Up to 19 H ₂ molecules can be adsorbed on the surface of Be‐decorated Li‐doped h‐BN, exhibiting an average adsorption energy ranging from −0.23 to −0.42 eV per H ₂ molecule. Electronic structure analysis reveals that the introduction of Li and Be atoms significantly alters the electronic properties of pristine h‐BN, as indicated by the modification of bands near the Fermi level, thereby enhancing conductivity and inducing semiconducting behavior. Furthermore, Be decoration Li doping collectively improves the hydrogen storage capacity (HSC), achieving a theoretical gravimetric density of up to 11.22 wt.%. The thermal stability of the H ₂ ‐adsorbed system was further evaluated using ab initio molecular dynamics (AIMD) simulations and the van’t Hoff equation, confirming its robustness under thermal conditions.

The growing adoption of flexible and portable electronic devices has created an increasing need for sustainable power sources that can operate without relying on conventional batteries. Triboelectric nanogenerators (TENGs) have attracted significant attention due to their ability to convert ambient mechanical energy into electrical energy. However, practical contact–separation TENGs may require additional spacers, elastic supports, or other structural components to maintain repeated contact and separation. This study presents a spacer-free arch-shaped TENG based on a laminated structure consisting of polytetrafluoroethylene (PTFE), paper, and aluminum. The arch configuration provides an inherent restoring tendency that facilitates contact–separation operation without a discrete spacer or additional elastic support. The fabricated device was experimentally evaluated through open-circuit voltage, short-circuit current, load-dependent electrical characterization, power measurement, capacitor charging, and LED illumination. Under the manual mechanical actuation conditions used in the present study, the device exhibited a maximum open-circuit voltage of 69 V, a reported peak short-circuit current of 68 $\mathsf{\mu }$A, and a peak power output of 8.576 $\mathsf{\mu }$W at an optimal load resistance of approximately 55 M$\Omega$. Furthermore, the TENG charged a 4.7 µF capacitor to 7 V within approximately 60 s and directly illuminated 30 commercial light-emitting diodes (LEDs). These results demonstrate the feasibility of combining a simple spacer-free arch-shaped configuration with readily available materials for low-cost triboelectric energy harvesting. Long-term cyclic durability, controlled mechanical excitation, environmental stability, and matched structural control experiments remain important areas for future investigation.