Need for off-grid electric vehicle (EV) charging solutions, using intelligent control systems, such as machine learning (ML)-based maximum power point tracking (MPPT), to harness solar energy, offers a means of optimizing efficiency even in the face of fluctuations. This innovative strategy combines clean energy, cutting-edge power electronics, and practical application, which makes it perfect for fostering innovation in areas with inadequate infrastructure. For rural areas without grid infrastructure, this paper presents a novel design and performance assessment of a portable solar-powered EV charging system. To maximize solar energy harvesting and charging efficiency, the system combines an interleaved buck-boost converter with an ML-driven MPPT algorithm. It is appropriate for small electric vehicles (EVs) like auto rickshaws because it uses a 48 V lithium iron phosphate (LiFePO₄) battery. A supervised regression model trained on real-time electrical (voltage, current, and power) and environmental (temperature, irradiance) parameters is used to implement the MPPT algorithm. The system was created using MATLAB/Simulink, and the key performance parameters were evaluated using real-time information. Analyses of the key performance metrics like charging efficiency, converter stability, and tracking accuracy show a superior energy harvesting efficiency of 97%.
This paper offers a structured methodology for regulating optimal control switching angles in multilevel inverters to achieve selective harmonic elimination (SHE) using embedded optimization algorithms on the PYNQ‑Z2 FPGA operating base. The proposed approach employs metaheuristic techniques including particle swarm optimization (PSO), genetic algorithm (GA), gray wolf optimization (GWO), slime mould algorithm (SMA), and whale optimization algorithm (WOA) to generate candidate switching angles across a wide range of modulation indices. For each modulation index, the most effective solution towards harmonic reduction and waveform quality is selected and implemented on the FPGA controller, enabling reliable real‑time operation with very low delay. Experimental validation of a 31‑level single‑phase inverter confirms the effectiveness of the method in eliminating selected harmonics and refining the fundamental output. Combining comparative algorithmic selection with FPGA‑based control demonstrates a systematic and efficient strategy for advanced inverter systems. A MATLAB/Simulink model is constructed to represent the 31-level inverter to enhance the practical aspect and study the frequency spectrum, where many harmonics that contribute to obtaining THD within the IEEE standards are eliminated. Also, the different types of losses are calculated based on the governing mathematical rates.
With the increasing global interest in renewable energy, advanced control strategies for wind energy conversion systems (WECSs) are critical for achieving not only maximum efficiency but also grid compatibility. This paper introduces a new adaptive, forecast-based predictive control algorithm to maximize the real-time energy efficiency of wind turbines integrated with smart grid architectures. The method couples a short-term wind-speed prediction module with a model-predictive control algorithm that uses the predictions to anticipate and adjust turbine operating conditions, such as pitch angle and rotor speed, to counteract the mechanical load from wind shear while maximizing power production. The analysis employs a dataset of 487 high-resolution cases of wind speed and wind turbine performance metrics to train and validate the adaptive model. The simulations were conducted in MATLAB/Simulink, using a model that captures the turbine’s nonlinear behavior and the electrical grid connection. The results indicate that the proposed controller increases the cycle-averaged power coefficient by approximately 8.2% and reduces tip-speed-ratio tracking error by roughly 70% relative to a conventional proportional-integral-derivative (PID) controller, while also reducing pitch and torque actuation, thereby lowering structural fatigue loads. This study is expected to provide a reliable solution to wind energy intermittency and facilitate the more stable and efficient integration of wind farms into modern smart grid systems.
The study herein investigates the structural development of cobalt(II) oxide (CoO) thin films grown on Ag (0 0 1) substrates and evaluates their impact on optical performance in white light-emitting diodes (WLEDs). The structural and chemical properties of CoO films were characterized using grazing-incidence X-ray diffraction (GIXD), Xray photoelectronic spectroscopic (XPS), as well as low-power electronic diffracting (LEED). The outcomes reveal that CoO films initially grow pseudomorphically up to 3-4 monolayers, followed by gradual relaxation toward the bulk lattice parameter. A periodic mismatch dislocation network forms at approximately 8 monolayers, leading to lattice tilting and the formation of ordered nanoscale domains. In addition, the incorporation of CoO as a scattering material in WLEDs significantly improves light scattering and color uniformity, while maintaining stable luminous performance. Compared with conventional scattering materials, CoO demonstrates lower scattering loss and reduced chromatic deviation, making it a promising candidate for high-precision optoelectronic applications.
This paper presents a simulation and comparison of direct power control (DPC) and the new backstepping (BS) control for a three-phase PWM rectifier. The objective is to evaluate these techniques in achieving high-performance grid integration, targeting the absorption of sinusoidal current, maintaining a unity power factor, and minimizing grid current harmonics. While DPC is recognized for its fast dynamic response, it often suffers from high power ripples and variable switching frequency due to the switching table. Conversely, the proposed BS control provides superior tracking performance and improved stability margin, even under parameter uncertainties. Through detailed modeling and simulation in MATLAB/ Simulink environment, this paper analyzes key performance indicators, including current total harmonic distortion (THD), response time, and robustness, under multiple grid conditions, and validates the proposed control law using a processor-in-the-loop (PIL) implementation. The comparative study demonstrates the trade-offs between the two control methods, highlighting the advantage of backstepping control.
This study presents a short-circuit and protection coordination analysis of a 33 kV/0.4 kV industrial distribution network supplying a phosphate processing plant in Mauritania. Using ETAP software, the system was modeled to evaluate fault behaviors based on IEC 60909 and IEC 60255-151 standards. The study assesses the existing protection architecture and proposes an optimized digital protection framework. Simulation results indicate that the initial symmetrical short-circuit current at the 0.4 kV busbar reaches 96.5 kA, exceeding the 25-36 kA breaking capacity of the legacy equipment by nearly 300%. The analysis also investigates single-line-to-ground faults, revealing a fault current inversion phenomenon driven by the solid grounding configuration. To address these vulnerabilities, the study proposes upgrading to 150 kA-rated high-breaking-capacity switchgear and implementing advanced negative-sequence protection (ANSI 46) and transformer differential protection (ANSI 87T). This work provides a practical diagnostic framework for mitigating asymmetrical faults and optimizing protection coordination in heavy-duty industrial clusters.
Wide-bandgap (WBG) semiconductor devices such as silicon carbide (SiC) and gallium nitride (GaN) enable higher switching frequencies, greater efficiency, and increased power density, but their fast-switching transients and steep dv/dt and di/dt characteristics generate substantially elevated electromagnetic interference (EMI). This review synthesizes peer-reviewed studies to quantitatively compare EMI mitigation outcomes across source-side modulation, gate-driving, filtering, packaging, and AI-based techniques for WBG converters. Packaging-integrated common-mode screens cut CM current by up to 26 dB while raising partial-discharge inception voltage by 53%. Chaotic PWM with passive filtering reaches up to 50 dB attenuation with a 74% reduction in inductor volume. AI-based closed-loop adaptation attains up to 19.2 dB average attenuation with 98.5% CISPR 25 compliance, while RL-based filters reach 25-30 dB across wide frequency ranges. Active gate-driving reduces peak EMI by 19-39 dB. No single technique simultaneously delivers the highest suppression, efficiency, and lowest cost. Hybrid Si/WBG design currently offers the most balanced trade-off. These outcomes are consolidated into a taxonomy of propagation mechanisms, mitigation techniques, application-specific strategies across five domains, and open gaps in standardized testing and validation. These findings provide a practical, quantitative reference for engineers designing next-generation, EMC-compliant WBG power electronic systems.
This study proposes the optimization of a hybrid genetic algorithm-particle swarm optimization (GA-PSO) building energy management combined with Six Sigma for quality control. The main problems include high energy consumption, large carbon emissions, and performance variability. Six Sigma is applied through control limits (UCL/LCL) and process capability index (Cpk) so that the solution is not only efficient but also stable. Using 30 days of operational data, the model evaluates daily energy consumption (kWh) and carbon emissions, then compares the baseline with pure GA, pure PSO, and GA-PSO+Six Sigma. The results show that GA-PSO reduces average energy consumption by 5.1% compared to GA and 3.5% compared to PSO. When combined with Six Sigma, the savings increased to 7.5% compared to GA and 6.6% compared to PSO, while reducing carbon emissions without compromising operational comfort. These findings present a measurable, sustainable, low-carbon building energy management model that is aligned with the decarbonization framework and ISO 50001 best practices.
To tackle the problem of modeling long-term trends and short-term and high-frequency variations in PV time series, a strong and efficient forecasting model of photovoltaic (PV) power generation is advanced. This paper presents STL-TCM-Former, a hybrid model that breaks down the raw PV signal with seasonal-trend decomposition with LOESS (STL) into trend and seasonal components. These elements are after that processed via a dual path encoder decoder transformer architecture that is improved with a temporal convolutional module (TCM). The period (transformer-based) path is used for capturing the global, long-range dependencies in seasonal component and temporal (TCM) path is used to extract the localized, short-term dynamics. Trend component is processed with a dedicated TCM branch, minimizing component interferences and providing a multiscale temporal representation, specific to PV generation patterns. The proposed model was evaluated on the Yulara (Uluru) solar dataset under short-term (60-hour) and long-term (300-hour) forecasting horizons. Compared with benchmark models including LSTM, WOA-LSTM, VMD-LSTM, WOA-VMD-LSTM, and hybrid WOA/VMD/LSTM configurations, STL-TCM-former achieved superior performance with R² = 99.76%, MAPE ≈ 2.24%, RMSE = 16.63, and MAE = 10.36. In addition, the PJM interconnection dataset was also employed to evaluate the generalization capability of the proposed method. The results demonstrate high accuracy, stability, and strong generalization capability under varying environmental conditions.
Three-phase active front-end (AFE) rectifiers are widely deployed in motor drives, electric vehicle chargers, and grid-connected renewable energy systems, where precise DC-link voltage regulation is essential for stable converter operation. In practice, DC-link capacitance degrades over time, and load profiles vary dynamically, both degrading the DC-link voltage regulation performance. Conventional proportional-integral (PI) outer voltage controllers are designed based on nominal operating conditions, with limited stability margins resulting in sluggish or oscillatory DC-link voltage responses under significant load and parameter variations. This paper proposes an adaptive internal model control-proportional integral (AIMC-PI) outer voltage loop controller for a three-phase AFE rectifier. It extends the conventional IMC-PI structure by incorporating an active damping term, an internal feedforward gain, a reference filter, and a Lyapunov-based adaptation law that updates the embedded plant model and IMC filter time constant online ensuring closed-loop stability and bounded tracking error. Simulation results show that AIMC-PI achieves a faster dynamic response than PI and performance comparable to IMC-PI and linear active disturbance rejection control (LADRC) under nominal conditions. As DC-link capacitance degrades to 0.5C, AIMC-PI maintains a well-damped DC-link voltage, whereas LADRC exhibits noticeable oscillations. Experimentally, AIMC-PI successfully eliminates the AC-supply current and DC-link voltage ripples present in fixed-λ IMC-PI.
The rapid growth of electric vehicles (EVs) requires intelligent and resilient planning of charging infrastructure under dynamic urban conditions. This paper proposes a city-scale spatiotemporal digital twin (SDT) that integrates LSTM-GNN fusion with resilience-driven hybrid optimization (GA-PSO-deep reinforcement learning) for adaptive EV infrastructure management. The LSTM model captures temporal variations in charging demand, while the graph neural network (GNN) learns spatial dependencies across charging stations, mobility networks, and grid components. Unlike existing approaches, the proposed framework incorporates power electronics-aware modeling, including charger power conversion system (PCS) efficiency, switching losses, and harmonic distortion constraints, ensuring realistic grid interaction. The digital twin also considers energy system metrics such as transformer loading, voltage deviation, and renewable energy variability, along with EV drive-cycle characteristics like fast charging and battery limits. Simulation results show that the proposed model improves demand prediction accuracy by 14-22%, reduces grid overload probability by 35%, lowers operational cost by 18%, and achieves improved power quality performance compared to conventional methods. The system maintains a high resilience index (>0.92) under stress scenarios. Overall, this work presents a holistic AI-driven digital twin framework that enhances grid stability, supports sustainable EV integration, and enables scalable deployment of future smart charging infrastructure.
Multilevel inverters (MLIs) are widely used in energy-conversion systems because they generate high-quality AC voltages with reduced harmonic distortion. However, existing MLIs often require numerous power devices, multiple DC sources, voltage sensors, or dedicated capacitor-balancing controllers, which increases both hardware cost and control complexity. This paper proposes a single-phase 15-level inverter derived from the Packed U-Cell structure. It employs a single DC source, three capacitors, and a reduced number of power switches. Based on the number of power switches, gate drivers, diodes, capacitors, DC sources, output levels, and total standing voltage per unit (TSVpu), the proposed topology achieves a lower cost function than the compared topologies reported in the literature. The proposed topology achieves a relatively low total standing voltage per unit (TSVpu) of 4.43. Sensorless open-loop SPWM offers inherent capacitor-voltage self-balancing, eliminating the need for voltage sensors or additional balancing loops. MATLAB/Simulink validation at 2 kHz and a modulation index of 1 covers steady-state operation, load transients, nonlinear loading, and DC-source voltage fluctuations. For a 50 Ω–20 mH load, the current THD is 2.20% without an output filter, confirming suitability for energy-conversion systems.
Electric vehicles (EVs) offer a sustainable mode of transportation; however, excessive battery temperature rise during charging degrades performance, limits lifespan, and affects the safe operation of power converters and EV drive systems. To address this problem, a back propagation neural network (BPNN) based temperature prediction model is integrated with four nature-inspired optimization techniques (NIOTs): whale optimization algorithm (WOA), moth flame optimization (MFO), modified particle swarm optimization (MPSO), and grey wolf optimization (GWO), to optimize multi-stage charging current profiles. Among the evaluated methods, WOA achieves the best performance, reducing charging time by approximately 10% (11400 s to 10300 s) and average temperature rise by nearly 50% (3.9 °C to 1.9 °C) compared to the conventional constant current-constant voltage (CC-CV) strategy. The optimized charging currents directly support improved DC-DC converter operation and stable EV drive performance by limiting thermal stress and current transients. Overall, accurate thermal-aware charging enhances charging efficiency, ensures safe converter operation, and contributes to reliable and long-life EV battery and drive system performance.
This work experimentally evaluates a simplified closed-loop on-off (hysteresis) controller for a single-phase buck converter using an LM324 comparator and an optocoupler-based gate drive to the IRF830 MOSFET. The controller compares the output voltage against a reference and drives the switch in a binary manner, yielding variable-frequency regulation with low component count. Performance is assessed for DC-DC regulation and dynamic reference tracking under sinusoidal, triangular, and rectangular waveforms at 1 Hz and 60 Hz, as well as under input-voltage droop, load perturbations, and real-time changes in reference waveform. Results show envelope-accurate tracking with fast transients and bounded ripple across all tested conditions, including representative grid-frequency operation. The controller maintains stability and preserves waveform fidelity despite supply and load disturbances; edge transients observed for rectangular references at 60 Hz are consistent with expected output-capacitor dynamics and remain bounded. The findings support the suitability of analogue on-off hysteresis with optocoupler isolation as a low-complexity approach for applications requiring efficient DC-DC conversion and dynamic signal tracking, especially where cost, simplicity, and rapid response are prioritized.
This paper presents a hybrid intelligent framework for fault diagnosis and localization in modern power distribution systems, addressing challenges such as noisy measurements, high-impedance faults (HIF), and uncertain operating conditions. The proposed approach integrates deep neural networks (DNN) for nonlinear feature extraction, support vector machines (SVM) for robust classification, and a fuzzy inference system for uncertainty-aware decision fusion, combining the strengths of deep learning, machine learning, and soft computing. A comprehensive dataset of over 12,000 fault instances is generated using IEEE 33-bus and 69-bus systems, covering multiple fault types (LG, LL, LLG, LLL), fault resistances (0.1-200 Ω), varying load conditions, and noise levels from 30 dB to -5 dB SNR. Wavelet-based denoising and hybrid feature extraction (time–frequency and statistical features) are employed to capture transient characteristics. The DNN generates discriminative feature embeddings, which are classified using an RBF-kernel SVM and further refined through fuzzy logic with Gaussian membership functions. Fault localization is performed using impedance-based estimation enhanced by learned correction. Results show that the proposed model achieves 98.5% classification accuracy, outperforming DNN (93.2%), SVM (90.4%), random forest (91.1%), and k-NN (88.6%). The model demonstrates strong noise robustness, with only -6% accuracy degradation at -5 dB SNR. It achieves fault localization error of 0.2-0.7 km and HIF detection with F1-score of 0.91. With inference latency of 45 ms (reduced to 28 ms after optimization), the system is suitable for real-time deployment, providing a scalable and reliable solution for smart grid fault monitoring.
The Cuk converter is widely used in power electronics due to its ability to provide both step-up and step-down voltage conversion with low ripple characteristics. However, conventional designs suffer from increased component count, higher losses, and significant voltage stress on switching devices. In this paper, an improved single-switch bridgeless AC-DC Ćuk converter is proposed for power factor correction (PFC) and loss reduction. The proposed topology employs a two-stage integrated structure with two low-value coupling capacitors (1 µF each). These capacitors effectively reduce voltage stress across the main switch and enhance switching performance. The operating principles of the proposed converter are analyzed under different switching conditions, highlighting its improved dynamic behavior and reduced component stress. Simulations using MATLAB/Simulink show that the proposed design achieves high efficiency, very low output voltage ripple (less than 1%), and a near-unity power factor of 0.9988. This performance, particularly the fast dynamic response, was achieved using a specially designed proportional-integral (PI) controller for output voltage regulation. The converter demonstrates fast dynamic response with a settling time of less than 0.23 s and a maximum overshoot of 7.5% under sudden load and input variations. Compared to conventional two-switch topologies, the proposed design significantly reduces switch voltage stress while maintaining a simpler structure, leading to lower cost and compact size. These results confirm that the proposed converter is a promising solution for high power quality applications.
Failures and guaranteed dependability of the electrical grid, early fault diagnosis in power transformers is essential. By examining gas ratios suggestive of faults, dissolved gas analysis (DGA) continues to be a vital component for transformer health monitoring. Using four preprocessing techniques raw data, min-max normalization, logarithmic transformation, and square root transformation; this study suggests a machine learning method for fault detection using DGA gas ratios (such as CH₄/H₂, C₂H₂/C₂H₄). Random forest (RF), support vector machines (SVM), gradient boosted trees (GBT), and a hybrid RF-GBT model that uses prediction fusion were the four supervised classifiers assessed. Performance was assessed using accuracy, precision, recall, f1-score, and Cohen's kappa. Experimental results show that the hybrid RF-GBT model with logarithmic transformation achieves the highest performance, with 94.93% accuracy and 92.37% Cohen's kappa, significantly outperforming individual classifiers. Data preprocessing, particularly logarithmic and square root transformations, enhances diagnostic robustness by mitigating feature skewness. This study underscores the importance of tailored preprocessing and ensemble methods for reliable transformer fault diagnosis.
Battery-powered electric vehicles (BEVs) are gaining significant attention due to high energy efficiency, zero emissions, and advanced control systems. This article illustrates the performance analysis of BEV, which consists of a high-voltage battery pack, a BEV controller, a motor driver, a gearbox, and a longitudinal driver. A conventional PID controller is used as a BEV controller. Several optimal algorithms are employed for tuning the PID controller, including the Ziegler-Nichols method (ZN method), particle swarm optimization (PSO), genetic algorithm (GA), grey wolf optimization (GWO), artificial bee colony algorithm (ABC), and artificial hummingbird algorithm (AHA). The proposed research framework was assessed in terms of vehicle efficiency, battery power consumption, vehicle mileage, motor speed, and battery state of charge (SOC). Metaheuristic algorithms with PID controllers outperform conventional PID and classical ZN-PID controllers. Among all algorithms, the PID-GWO controller achieves maximum vehicle mileage, low battery power consumption, improved battery SOC, and the highest vehicle efficiency.
This paper presents a hybrid electric vehicle charging station powered by both a PV source and the utility grid, incorporating an energy management strategy that prioritizes the utilization of solar energy while exporting surplus power to the grid during periods of low charging demand. To enhance the performance of maximum power point tracking, a hybrid control strategy integrating the grey wolf optimizer (GWO) and the super-twisting algorithm (STA) is proposed. The GWO performs rapid global exploration to accurately identify the maximum power point, whereas the STA ensures precise, robust, and chattering-free tracking under steady-state operating conditions. The proposed system was modeled in MATLAB/Simulink and validated under a dynamic irradiance profile characterized by both abrupt and gradual variations. Simulation results demonstrate a convergence time of 2-3 ms, residual power oscillations below 0.1%, and an average tracking efficiency of 99.38%. Compared with conventional MPPT techniques, the proposed STA-GWO approach significantly suppresses steady-state oscillations, accelerates convergence, and prevents MPP tracking failure under rapid irradiance fluctuations through the global optimization capability of GWO. These findings highlight the effectiveness of the proposed hybrid MPPT strategy in improving the robustness, energy conversion efficiency, and grid integration capability of PV-powered EV charging stations, making it a promising solution for next-generation sustainable charging infrastructure.
The permanent magnet synchronous motor (PMSM) is commonly used in industrial and home appliances for its high efficiency and dynamic performance. In this research, a PMSM drive based on field-oriented control (FOC) is designed and simulated using MATLAB/Simulink. The speed controller of the drive is tuned using the trial-and-error method. However, the method requires more time for testing and adjustment of the speed controller to generate an optimal output response. Thus, particle swarm optimization (PSO) and artificial bee colony (ABC) algorithms, which require less computational effort and effectively produce good responses, are used to optimize the speed controller of the drive. PSO and ABC also offer an attractive optimization framework because of their independent, agnostic model structures, global searching capability, and low-load real-time calculation. In this study, the results obtained from different tuning methods are compared to determine the best optimization method of the speed controller in the PMSM drive under different operations. Other than that, performance measures such as integral square error (ISE), integral absolute error (IAE), and integral time absolute error (ITAE), which are commonly used to measure the effectiveness of a controller, are also discussed. The results show that the PMSM drive based on the ITAE criterion using the ABC algorithm gives better performance under different conditions.
