A Comparative Study of Perturb & Observe and ANN-Based MPPT Algorithms Under Various Environmental Conditions

Abstract

This Photovoltaic energy is a promising renewable energy source because it provides a cleaner alternative to fossil fuels. However, the output power of a PV system is affected by factors like solar irradiance and temperature that can change and influence its performance. Therefore, techniques for maximum power point tracking must be developed so the photovoltaic system produces the maximum power. These techniques guarantee that the PV system has the highest power point at all times regardless of weather conditions. The two types of controllers that are presented in this paper are based on the traditional Perturbation and Observation method and the use of artificial neural networks. From simplicity, low cost, and suitability for medium to large photovoltaic systems, the Perturbation and Observation method is chosen. On the other hand, artificial neural networks is well suited to manage complex systems and may assist in improving maximum power point tracking. In addition, the performance of these two methods is compared to a fuzzy logic-based MPPT approach that was developed and published in a previous study. The controllers were evaluated under various environmental conditions using MATLAB/Simulink. The results demonstrated that the artificial neural networks-based controller outperformed both the Perturbation and Observation and fuzzy logic controller methods in terms of efficiency and overall performance. Additionally, the artificial neural networks approach significantly minimized power fluctuations. However, the fuzzy logic controller method showed a faster response in reaching the maximum power point compared to both the artificial neural networks and Perturbation and Observation techniques.