Comparative analysis of maximum power point tracking through ensemble learning and deep neural networks

This paper presents a maximum power point tracking (MPPT) methodology utilizing machine learning (ML) to precisely monitor the maximum power output of a photovoltaic (PV) panel. This approach demonstrates rapid and accurate tracking of the maximum power point (MPP) even under varying weather conditions, including fluctuations in solar irradiation and temperature. The objective of this study is to conduct a comparative performance analysis of three ML-based MPPT algorithms, namely Random Forest, Feedforward Neural Network, and Recurrent Neural Network (RNN). Among the three algorithms, Random Forest exhibited superior performance with reduced execution time.