Output list
1–10 of 100 results
Journal article
Published 2026
IEEE transactions on transportation electrification, Early Access
Charging/discharging power data for an electric vehicle (EV) customer with a vehicle-to-grid (V2G) contract is very privacy-sensitive. To address this privacy concern, the V2G-based battery dispatch optimization (BDO) based on the alternating direction method of multipliers with input-discriminative local differential privacy (ID-LDP-based ADMM) has been developed. This ID-LDP-based ADMM algorithm involves a master problem and sub-problems, where each sub-problem represents the V2G-based battery dispatch optimization dispatch of an individual EV. Moreover, the input-discriminative local differential privacy budget of EVs can be dynamically adjusted to enhance privacy protection during critical time periods against potential attacks to the untrusted data curator in the master problem. Meanwhile, it can be proven that the optimal charge-discharge power solutions of EVs are not interfered with input-discriminative local differential privacy noise, eliminating the need for a trustworthy data curator. However, the existing DP-ADMM algorithm has to face practical challenges in balancing privacy levels, accuracy, and convergence speed within a trustworthy data curator. Case studies demonstrate that the ID-LDP-based ADMM algorithm converges to the optimal solution in about 10.6% of the time required by existing methods, even with different privacy protection levels, i.e., up to 65% larger obfuscation range than existing methods during critical time periods.
Journal article
Published 2026
Sensors (Basel, Switzerland), 26, 17, 5596
Cavitation is one of the major factors reducing the performance, efficiency, and service life of deep well submersible pumps used in agricultural irrigation systems. Therefore, the early detection of cavitation is essential for improving pump reliability and reducing maintenance costs. In this study, a machine learning-based framework is proposed to detect and classify cavitation conditions using experimental data collected from a deep well pump test unit. Hydraulic and operational parameters were measured under different operating conditions, while the measured noise level was used only to assign cavitation labels during dataset preparation. According to the measured noise level, the operating conditions were classified into three categories: Normal, Incipient Cavitation, and Severe Cavitation. Several machine learning algorithms were evaluated using stratified cross-validation and an independent test dataset. Model performance was assessed using Accuracy, Precision, Recall, and F1-score. The results showed that the Extra Tree classifier achieved the best performance with an accuracy of 82.4%. Feature importance analysis indicated that power consumption and submergence depth were the most influential parameters for cavitation detection. Unlike many existing studies that rely on computationally intensive models, the proposed framework employs a simple and lightweight machine learning approach while maintaining reliable prediction performance. Its low computational complexity makes it a promising candidate for future implementation on resource-constrained edge devices, enabling real-time cavitation monitoring in agricultural pumping systems.
Journal article
Published 2026
Energy reports, 16, 109721
This study aims to ensure the continuity of supplying the critical loads in electrical power systems through mobile energy storage systems (ESSs) powered by renewable energy sources, and to benchmark their performance against stationary alternatives. An integrated planning framework is developed to couple the electric network with road-transport logistics, extending conventional lifecycle costing to include both electrical network infrastructure and the costs of battery mobility. To enable realistic planning and operational analysis, a multi-regional power system planning and operation model and a battery transport and logistics model is jointly formulated and solved to represent the behavior of mobile and stationary ESSs within the power system. The empirical focus is on a real-world example of northwest and western Iran, evaluating mobile photovoltaic supported ESS setups versus stationary configurations for sustaining 50–100 kW critical loads during grid outages. The assessment accounts for seasonal resource variability, deployment windows, setup times, and inverter/battery constraints across four scenarios, and complements these analyses with projections for fixed solar farms in 2030, 2040, and 2050 under rising solar penetration. Results indicate that mobile systems deliver rapid, relocatable security for critical loads, whereas stationary assets provide scalable, planned robustness over longer horizons; together, they constitute a complementary pathway to reliable, low-carbon electricity supply.
Journal article
Published 2026
Energy engineering, 123, 10, 12
This paper presents a probabilistic optimization framework for enhancing the photovoltaic hosting capacity of distribution networks through the coordinated operation of smart photovoltaic-static synchronous com-pensator (PV-STATCOM) inverters. High penetration of photovoltaic systems introduces significant operational challenges, including voltage rise, increased power losses, and reduced system reliability, particularly under uncertain load demand and intermittent solar generation. To address these challenges, a multi-objective optimization problem is formulated to simultaneously maximize the photovoltaic hosting capacity while minimizing voltage deviation and power losses, subject to network operational constraints. The proposed approach employs the Fire Hawk optimizer, a recently developed metaheuristic algorithm, to efficiently solve this nonlinear and non-convex optimization problem. The decision variables include the optimal sizing of photovoltaic systems and the reactive power support provided by PV-STATCOM inverters. Unlike conventional deterministic approaches, uncertainty in load demand is explicitly modeled using a probabilistic framework based on Latin Hypercube sampling and Monte Carlo simulation. The Latin Hypercube sampling is utilized for efficient scenario generation while Monte Carlo studies enable statistical evaluation of system performance under diverse operating conditions. The effectiveness of the proposed method is validated on the IEEE 15-bus distribution system. Simulation results demonstrate that the coordinated control of PV-STATCOM inverters significantly improves voltage profiles and reduces power losses, leading to an increase in photovoltaic hosting capacity from 6.89 to 8.311 MW, corresponding to a 20.62% enhancement in the network under study. Furthermore, studies demonstrate that the Fire Hawk optimizer algorithm exhibits fast convergence and superior performance compared to conventional optimization techniques, including genetic algorithm, particle swarm optimization, and Aquila optimizer. The probabilistic analysis also confirms improved system robustness by reducing voltage violation risks under uncertainty.
Journal article
Hybrid CNN-LSTM modeling with attention mechanism for solar power forecasting
Published 2026
Discover applied sciences
Accurate short-term forecasting of solar photovoltaic (PV) power is essential for improving power-electronics-based control and energy management in modern renewable energy systems. This study proposes a hybrid deep learning framework that integrates convolutional neural networks, long short-term memory units, and an attention mechanism to predict hourly PV power with high precision, enabling dynamic operation of dc-dc converters, inverters, and grid-connected PV units. To address the challenge of missing meteorological data in the first considered site, a Random Forest regression model is employed to reconstruct the incomplete solar radiation and cloudiness data. The framework was then pretrained on the PV data of a second site (Alice Springs, 2014–2017) and then fine-tuned using a five-month dataset collected from a PV plant near Urmia, Iran (Sep 2024–Feb 2025). Through numerical studies, it is shown that the attention layer of the proposed technique further enhances interpretability by highlighting time intervals that are most influential for power conditioning and converter response. Notably, training exclusively on the first site data outperformed the transfer learning setup, achieving a coefficient of determination (R 2) of 0.9500, root mean squared error (RMSE) of 0.0688, mean absolute error (MAE) of 0.0321, and a peak error of 0.0398% under high-generation conditions. This level of performance supports more reliable short-term adjustment of converter setpoints and inverter dispatch, contributing directly to voltage stability and efficient energy management in grid-connected PV systems. Overall, the proposed approach reinforces the synergy between artificial intelligence driven forecasting and power electronics control, offering a scalable and interpretable solution that supports renewable energy integration and smart grid stability, particularly in data-limited environments.
Journal article
Published 2026
Machine learning with applications, 25, 100962
Electric vehicle proliferation necessitates efficient power conversion systems with minimal harmonic distortion. Among the established solutions for electric vehicles’ charging infrastructure, 12-pulse thyristor-based controlled rectifiers integrated with phase-shifting transformers offer superior harmonic performance; however, their advanced design requires iterative refinements to address nonlinear magnetic behavior, loss mechanisms, and complex trade-offs among harmonic mitigation, transformer rating, volume, cost, and thermal limits. This study presents an explainable artificial intelligence (XAI)-enhanced framework that moves beyond simulation-driven, trial-and-error design toward informed data-driven decision-making supported by interpretable feature-importance analysis. The proposed framework transforms black-box ensemble models into transparent decision-support tools, enabling engineers to better understand the influence of key design parameters on grid-current total harmonic distortion (THD). An ANSYS TwinBuilder–Maxwell 3D co-simulation is employed to generate 4750 operating points linking the thyristor firing angle, output filter capacitance, and load resistance to the grid-current THD for training 15 machine learning algorithms. The evaluated machine learning models demonstrate that tree-based ensemble methods provide the highest predictive performance, with CatBoost achieving R² = 0.964 and RMSE = 0.358. SHAP-based explainability analysis identifies the firing angle as the dominant control parameter (54% feature contribution), followed by load resistance (34%) and filter capacitance (12%). Validation against an independent ANSYS Maxwell–Twin Builder co-simulation at a representative operating point resulted in a prediction error of 1.18%, demonstrating good agreement between the surrogate model and the simulation within the investigated parameter space.
Journal article
Published 2026
Energy conversion and management. X, 31, 102174
Regulation plays a crucial role in encouraging distribution system operators (DSOs) to enhance the technical performance of power distribution systems. Meanwhile, the increasing deployment of fast charging stations (FCSs) offers considerable flexibility that can be exploited to enhance system reliability and peak-shaving capability through appropriately designed incentives. To coordinate the interactions among the regulator, DSO, and FCS aggregator, this paper proposes a regulatory-driven planning framework for optimal FCS deployment by integrating both the FCS planning model and the DSO’s operational problem into a unified decision-making framework. Within this framework, the regulator incentivizes the DSO to enhance system reliability under N−1 contingencies and peak-shaving performance under normal state through coordinated vehicle-to-grid and grid-to-vehicle services provided by the aggregator. Accordingly, a four-step planning process is developed. First, the baseline FCS planning is determined by the aggregator. Next, the regulator establishes a reliability-oriented penalty-reward model and a peak-shaving-focused reward model. Subsequently, the baseline objective functions of the DSO and aggregator are identified to evaluate their economic interactions. Finally, a system-oriented problem jointly optimizes the FCS planning model and the DSO’s operational problem under uncertainties in traffic flow and consumers’ demand. Case studies on the IEEE 33 and 69-bus test systems demonstrate the effectiveness and scalability of the proposed framework. Compared with conventional baseline models, the proposed framework increases the economic profits of the DSO and aggregator respectively by 37.0 and 56.0%, for the IEEE 33-bus system and respectively by 50.4 and 22.58%, for the IEEE 69-bus system. Furthermore, it enhances system reliability respectively by 39.9 and 62.53% for the IEEE 33 and 69-bus test systems, while improving peak-shaving capability by 25.0% for both systems. The results demonstrate that the proposed framework provides a practical regulatory mechanism for economically efficient FCS deployment while enhancing distribution system performance and creating a win-win outcome for both the DSO and aggregator.
[Display omitted]
Journal article
Published 2026
Energies (Basel), 19, 13, 3065
Integrating solar power into electricity grids requires accurate short-term forecasting of the global horizontal irradiance to accurately predict the expected solar power generation. This paper compares five tree-based machine learning models against a Persistence baseline for multi-resolution forecasting in arid climates. A 13-year dataset from Basra, Iraq, has been employed in this study for verification purposes, and the models are tested across various very-short- to short-term forecasting horizons of 5, 10, 15, 30, and 60 min. Unlike most existing studies that focus on single forecasting horizons or mixed climatic conditions, this work systematically benchmarks multi-resolution irradiance forecasting under distinct sky conditions in a hot arid environment using a strict anti-data-leakage framework. To avoid data leakage in these models, feature engineering has used only lagged inputs. The dataset has been split into three groups for training, validation, and testing (respectively 70, 15, and 15% of the entire available dataset). The models were then tested separately under clear, partly cloudy, and cloudy skies. Numerical studies prove that picking the best model depends heavily on the forecast horizon. For very-short-term predictions, the Persistence model was competitive (RMSE = 21.32 W/m2), while the Gradient Boosting model proved slightly more accurate (RMSE = 17.65 W/m2). For the 60 min horizon, the boosting models took a clear lead. The HistGradientBoosting model resulted in a 67% reduction in the RMSE compared to the Persistence baseline. Also, the top-performing model changed depending on the weather and the time scale. Gradient Boosting was the clear winner for short-term clear sky forecasts, while XGBoost handled the longer horizons. Partly cloudy skies showed a rotating mix of different boosting algorithms taking the lead. However, studies show that when skies were fully overcast, complex machine learning models fail to capture chaotic patterns, making the simple Persistence baseline a necessary reliability safeguard. The results reveal that no single model consistently dominates all forecasting horizons and weather conditions, highlighting the necessity of adaptive model selection for operational solar forecasting. These findings highlight the importance of horizon- and weather-adaptive model selection for operational solar forecasting. Rather than relying on a single universal algorithm, grid operators in arid regions can improve forecasting reliability by dynamically selecting models based on prevailing sky conditions and forecast horizons.
Journal article
Published 2026
Energy conversion and management. X, 31, 102022
The decarbonization of transport and energy sectors requires rapid deployment of electric vehicle (EV) fast charging infrastructure integrated with distributed energy resources (DERs). Although joint planning of fast charging stations (FCSs), photovoltaics (PV), and energy storage systems (ESSs) has been widely studied, many existing frameworks rely on simplified operating assumptions and do not capture temperature-driven variations in asset performance or weather-load interactions. This paper proposes a two-stage stochastic planning framework that explicitly incorporates ambient temperature effects and aligns investment decisions with regulatory incentives. The model embeds temperature-dependent physical constraints for EV energy consumption, charging power limits, EV/ESS charging efficiency, ESS discharging efficiency, and ESS degradation. Two performance-based incentive mechanisms are incorporated to monetize system-supportive behavior, namely self-sufficiency and peak shaving. Uncertainty is modeled using a vine copula scenario generation method that preserves nonlinear and tail dependencies among solar irradiance, ambient temperature, and electrical load. Case studies on a modified IEEE 33-bus system across three climatic zones demonstrate the importance of temperature-aware planning. Under cold climate conditions, traditional planning models underestimate required FCS capacity by up to 67% and lead to substantial fleet energy deficits under realistic operation conditions. Repeating the study over a larger IEEE 69-bus system with 50% higher EV penetration verifies the scalability of the proposed technique and confirms the findings that even mild climate locations may experience significant fleet energy deficits when temperature effects are ignored, and that unserved energy may occur under extreme cold. The proposed framework yields investment plans that maintain operational adequacy and improve economic performance through effective incentive capture.
[Display omitted]
•A temperature-aware stochastic planning framework for FCSs with DERs is proposed.•Regulatory reward models monetize self-sufficiency and peak shaving.•A vine copula captures dependencies among weather variables and electrical load.•Traditional models significantly underestimate FCS needs in cold climates.
Journal article
Published 2026
Energies (Basel), 19, 12, 2905
The trend in the power electronics industry toward higher power density and efficiency has brought high-frequency transformers (HFTs) to the forefront of critical applications, including isolated DC–DC converters, electric vehicle chargers, and solid-state transformers. This paper focuses on the leakage inductance of HFTs and presents a systematic comparative framework that evaluates five surrogate modeling and hybrid optimization approaches for the rapid and accurate estimation of leakage inductance. A comprehensive parametric dataset was constructed, comprising 1210 finite element analysis simulations conducted via finite element analysis in the ANSYS Maxwell 2024 R1 environment, varying the number of winding turns, primary winding thickness, and secondary winding thickness of the HFT. All five methods were trained and evaluated on the same dataset under identical conditions. The comparative evaluation demonstrates that the proposed hybrid Gray Wolf optimizer–artificial neural network (GWO-ANN) framework achieved the highest prediction accuracy (R2 = 0.9832, MSE = 0.01780, MAE = 0.0935 µH) and the fastest convergence among all tested approaches. The generalization capability of the proposed model was confirmed through blind validation tests across six geometric configurations spanning the full range of the design space, yielding a maximum prediction error of 8.15% and an average error of 2.14%. The functional validity of the proposed parameters was further tested in a third validation layer using MATLAB/Simulink R2024b transformer circuit studies, demonstrating a theoretical efficiency of 96.06%. This three-layer validation approach proves both the parametric and functional reliability of the proposed framework for HFT designs.