Accuracy of Detection and Classification of DC Faults using Levenberg Marquardt Based Back Propagation Algorithm

Abstract

Vitality is seen as a prime administrator in the time of wealth and a vital figure budgetary headway. Obliged fossil resources and natural issues associated with them have underscored the necessity for new reasonable vitality supply choices that uses sustainable power sources. Among open developments for essentialness age from solar dependent sources, the photovoltaic system might give a gigantic pledge to develop a progressively feasible imperativeness structure. This paper presents accuracy of detecting DC faults in a photovoltaic (PV) framework based on Levenberg - Marquardt (LM) neural network. The result showed that this model based on LM neural network is effectual to grip doubts and nonlinearities of DC side faults in PV module without using mathematical model. All probable faults in DC side of PV system are obtained over 100 kW plant. It is discovered that the proposed framework has demonstrated its integrity for the pragmatic applications.