Year: 2025
Type: article
Abstract: The study presents an overview and a simulation of maximum power point tracking (MPPT) for Photovoltaic (PV) systems that uses an artificial neural network (ANN) controller as proof of concept. Solar energy must be harvested with high efficien-cy as the world turns to renewables. The usual Perturb and Observe (P&O) and Incremental (InC) method loses power by oscil-lating around the Maximum Power Point (MPP) and reacts slowly to sudden weather changes. The work therefore tests an ANN as a better choice. The authors survey earlier ANN MPPT studies that cover many network types, training schemes and mixed strategies. They then build a MATLAB/Simulink model that runs an ANN controller and a P&O controller on the same PV array. The ANN learns from Istanbul 2020 weather data. The results show the ANN reaches 252 W and 87.9% of efficiency while P&O reaches 241 W and 84.26% of efficiency, and InC reaches 245 W and 78.1% of efficiency. The ANN also tracks the MPP faster and with steadier behaviour when irradiance varies. These outcomes confirm that ANN MPPT can raise the energy output of PV systems.
Authors: Mehdi Rezaee, Yusuf Gürcan Şahin
Institution: İstanbul Gelişim Üniversitesi
Language: English
Field: Energy
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4412374482