Installation of RBF network using artificial intelligence for the purpose of photovoltaic fault detection

Abstract

This research presents an ANN-based defect detection method for use in solar systems. This study introduces a new approach to PV defect detection using ANN. Despite the abundance of literature in the subject, the training, validation, and testing of the Radial Basis Function (RBF) network only need two inputs. The network achieved an unparalleled detection accuracy of 98.1%. Before training the network, the data set is subjected to the suggested methodology's "mapping of inputs" technique, which goes beyond data normalisation. Testing the network in partially shaded and cloudy situations further supports its accuracy.