A Comparative Study of the Proposed ANN-based Machine Learning MPPT Method in a Large-Scale Grid-Connected PV System under Variable Climatic Conditions

In response to the escalating demand for solar energy, driven by the depletion of fossil resources and their environmental repercussions, numerous studies have been undertaken to enhance the tracking of the maximum power point, optimizing the utilization of solar cell power. Controllers such as P&O, INC, and Fuzzy Logic have been subjects of investigation. This article explores the performance of proposed controllers, employing the extraction and integration of valuable data from these controllers. By synthesizing and training the data derived from various methods, a more effective approach, emphasizing speed and accuracy, has been devised using artificial neural network and machine learning tools. The effectiveness of this technique has been verified through simulations conducted on a large-scale grid-connected PV system within the MATLAB/Simulink, considering various atmospheric conditions.