ANN Training using Fireworks Algorithm and its Variants for PV Array Fault Classification

In recent years, strong photovoltaic systems are now more necessary than ever, and this development has been accompanied by the demand for diagnostic systems characterized by high accuracy and robustness, whether in normal operation or in the face of unexpected events. Recently, the development of an accurate, stable, and fast intelligent diagnostic system has been the main focus of research. In this article, the fireworks algorithm, dynamic search in fireworks algorithm, and adaptive fireworks algorithm are used to train the artificial neural network to enhance its accuracy and convergence time performance. The combination aims to compare the fireworks algorithm and its variants to determine the most efficient version concerning artificial neural network optimization in order to produce a fast, robust and intelligent photovoltaic fault classification model.