Performance analysis of evolutionary based optimization techniques for gmppt of photovoltaic systems

Abstract

Ambient temperature, solar irradiance, and dynamic environmental newlineconditions significantly influence the performance of solar photovoltaic (PV) newlinepanels. In partial shading conditions (PSC), multiple peaks appear in the I-V newlineand P-V curves, leading to substantial power losses. Conventional maximum newlinepower point tracking (MPPT) methods struggle to efficiently track the global newlinemaximum power point (GMPP) in these conditions. To address this newlinelimitation, this thesis proposes a novel hybrid MPPT strategy combining the newlineGrasshopper Optimization Algorithm (GOA) and Support Vector Machine newline(SVM) to enhance tracking accuracy and efficiency. Unlike traditional newlineheuristic-based MPPT methods, which may suffer from slow convergence newlineand local optima trapping, the GOA-SVM hybrid model leverages the newlineexploratory power of GOA and the predictive capabilities of SVM to newlinedynamically adjust to environmental fluctuations. newlineThe proposed approach has been validated in both MATLAB/SIMULINK newlinesimulations and real-time hardware experiments using a boost converter and newlinediverse PV array patterns. Simulation results demonstrate significant newlineimprovements in convergence speed, accuracy, and reduced oscillations newlinearound the GMPP compared to conventional algorithms. Hardware newlineimplementation further confirms the feasibility of the proposed method in newlinepractical scenarios. One of the key contributions is the integration of a selfadaptive newlinefeedback loop, where SVM continuously updates its predictions newlinebased on real-time environmental changes, ensuring sustained efficiency newlineunder varying conditions. newline