A Framework for Fault Recognition of Photovoltaic Plants with Convolutional Neural Network

Emerging innovations in the usage of photovoltaic (PV) power are mostly attributable to new global regulations aimed at limiting the consumption of fossil fuels. In addition to being impacted by many types of failures, atmospheric factors have a significant impact on the effectiveness of PV systems, which can result in significant dissipation of energy during control system. This study suggests a unique approach for locating defects in photovoltaic panels that makes use of a high-resolution camera mounted on a vehicle. These devices create a lot of information which needs to be handled and examined. In this research, a new method for locating panels that can detect fault is presented. For this, photovoltaic image dataset is utilized. Then, pre-processing is done to eliminate the noise and normalize the images. For the purpose of creating a reliable detection framework, a unique convolutional neural network is combined. During the examination, fault diagnosis is made possible by the automated processing of the data that are collected. By validating the performance metrics of proposed method with other techniques, it performs well than others. The findings reveal an overall accuracy of 99.43% for the developed technique.