Résumé: hotovoltaic panels represent the most abundant source of renewable energy and the cleanest form of electrical energy derived from the sun. However, partial shading can lead to the appearance of multiple local maximum power points (LMPP) in the power-voltage (P-V) characteristics of solar panels. This situation traps classical power maximization algorithms, such as perturb and observe (P&O) or incremental conductance, as these algorithms tend to deviate from the global maximum power point (GMPP), resulting in reduced electrical energy production. To overcome this major challenge in the electrical industry, we propose in this study a hybrid grey wolf optimization-perturb and observe hybrid (GWO-P&O) algorithm, designed to converge towards the global maximum power without being trapped in local peaks. To demonstrate its effectiveness, the proposed algorithm was simulated in MATLAB/Simulink under various complex and uniform partial shading conditions. Furthermore, a comparative study was conducted with the P&O and GWO algorithms to evaluate precision, tracking, response time, and efficiency. The simulation results revealed superior performance for the proposed technique, particularly in terms of constant tracking of the global peak, with efficiencies of 99.95% and 99.98% in the best cases, faster response times (ranging from 0.07 to 0.04 s), and minimal, almost negligible oscillations around the GMPP.

Résumé: Photovoltaic (PV) systems have emerged as a key area in renewable energy research, propelled by increasing global energy demand and the necessity for sustainable alternatives. However, the inherent nonlinearity of PV arrays, coupled with their sensitivity to fluctuating environmental conditions such as solar irradiance and temperature, often limits their ability to operate at maximum power output. This paper presents a novel approach for maximizing the power extraction from PV systems through the intelligent integration of Artificial Neural Networks (ANNs) within the MATLAB/Simulink environment. Traditional Maximum Power Point Tracking (MPPT) techniques often struggle with dynamic changes and may exhibit oscillations around the optimal point. Utilizing the learning and adaptive capabilities of ANNs, a robust and efficient MPPT controller is developed. The ANN is trained offline using a comprehensive dataset including various operating conditions, enabling it to accurately predict the maximum power point (MPP) voltage or current. The proposed ANN-based MPPT algorithm is then effectively applied and rigorously evaluated in MATLAB/Simulink, demonstrating superior tracking accuracy, faster convergence, and reduced power losses compared to conventional methods under rapidly changing atmospheric conditions. This research highlights the significant potential of ANNs to enhance the performance and reliability of PV systems, facilitating more efficient and cost-effective solar energy utilization

Résumé: This study focuses on the modeling and simulation of a photovoltaic (PV) energy conversion system that incorporates a Boost DC-DC converter along with multiple Maximum Power Point Tracking (MPPT) algorithms. The primary objective is to enhance solar energy harvesting efficiency under fluctuating irradiance and temperature conditions. Three MPPT strategies are examined and compared: the conventional Perturb and Observe (P&O) method, the Incremental Conductance (INC) algorithm, and a Neural Network (NN)-based technique. The PV system and control strategies are simulated using MATLAB/Simulink. Performance evaluation criteria include tracking accuracy, convergence speed, and response under dynamic environmental changes. The results show that while the P&O and INC algorithms offer simplicity and acceptable performance in steady-state conditions, the neural network approach provides faster tracking and better accuracy during rapid variations in solar irradiance. These findings highlight the potential of artificial intelligence in improving the efficiency and adaptability of PV energy systems.

Résumé: This study examines the effect of the improved controller on the load frequency control (LFC) issue. The proportional-integral-derivative (PID) parameters are calculated for a single-area control system utilizing genetic algorithms (GA) and particle swarm optimization (PSO). The LFC is a stochastic issue resulting from load fluctuations and alterations in system operating conditions. This leads to the conventional controller’s inability to modify the Load Frequency Control when employing a traditional PID.To implement optimal PID controller parameters with Genetic Algorithms and Particle Swarm Optimization. The findings demonstrate the precision and resilience of executing the optimized controller parameters. The proposed optimized controller has superior control quality compared to the standard controller. MATLAB/Simulink is employed to resolve the system equations. The proposed optimized controller demonstrates rapid reaction and superior control quality compared to the standard controller

Résumé: Optimizing energy extraction from photovoltaic (PV) systems necessitates effective maximum power point tracking (MPPT) solutions owing to the nonlinear properties of PV modules and the fluctuation of ambient circumstances. Traditional algorithms like Perturb and Observe (P&O) are prevalent due to their simplicity; yet, they frequently demonstrate steady-state oscillations and diminished accuracy in conditions of rapidly fluctuating irradiance. In contrast, Artificial Neural Networks (ANN) provide adaptive and nonlinear modeling capabilities, facilitating more accurate MPP prediction. This study conducts a comparative comparison of ANN-based and P&O MPPT approaches for a boost-converted photovoltaic system utilizing MATLAB/Simulink simulations. The artificial neural network (ANN) is executed as a Multi-Layer Perceptron (MLP) trained on photovoltaic current-voltage datasets under diverse irradiance and temperature conditions, whereas the Perturb and Observe (P&O) method depends on iterative duty-cycle perturbation. Simulation results indicate that the ANN-based MPPT attains quicker convergence, diminished oscillations, and superior tracking efficiency relative to the traditional P&O method, especially in dynamic weather circumstances. These findings validate the efficacy of ANN-based controllers in enhancing PV system performance..

Résumé: This study describes how to model and simulate a photovoltaic (PV) system that is connected to a three-phase power grid through a single static converter. Renewable energy sources must be added to modern power networks in a way that is both cost-effective and efficient. Using a single converter makes the system less complicated and guarantees dependable operation. The photovoltaic generator has nonlinear current-voltage characteristics, and a maximum power point tracking (MPPT) algorithm is used to make sure that the best amount of energy is extracted even when the temperature and irradiance change. Sinusoidal pulse width modulation (SPWM) controls the grid-connected inverter. This makes it easier to sync with the three-phase grid and makes sure that the power injection quality is high. The MATLAB/Simulink simulation looks at important performance measures such output voltage, injected current and power factor. The results confirm that the proposed configuration delivers stable performance with enhanced conversion efficiency. The study demonstrates the feasibility of integrating solar devices into the grid through an optimized framework and highlights prospects for further experimental validation.

Résumé: This work focuses on the modeling, simulation, and comparative analysis of a photovoltaic (PV) energy conversion system equipped with a DC-DC Boost converter and controlled by four different Maximum Power Point Tracking (MPPT) algorithms: the conventional Perturb and Observe (P&O), Improved P&O, Incremental Conductance (INC), and Improved INC methods. The goal of this work is to compare the performance of each method in terms of dynamic response, steady-state accuracy, and tracking efficiency under varying operating conditions of solar irradiance and temperature. The PV system is modeled in the MATLAB/Simulink environment using a realistic photovoltaic module connected to a Boost converter feeding a resistive load. The converter duty cycle is adjusted by the MPPT controller to keep the PV array operating at the MPP

Résumé: This paper investigates the synergistic integration of the Perturb and Observe (P&O) Maximum Power Point Tracking (MPPT) algorithm with lithium-ion battery storage in photovoltaic (PV) systems to achieve enhanced energy harvesting and utilization. The P&O MPPT technique is employed to dynamically track the maximum power point of the PV array under varying irradiance and temperature conditions, maximizing energy extraction. The integrated lithium-ion battery system provides efficient energy storage, enabling the decoupling of energy generation and demand, improving system autonomy, and mitigating the intermittency of solar energy. This paper details the operational principles of the combined system, analyzes its performance characteristics through simulation and highlights the benefits of this integrated approach in terms of increased energy yield, improved system stability. A bidirectional Buck Boost converter Controls the charging and discharging of the battery, regulating the voltage and current flow to protect the battery from overcharge and deep discharge.