Location for fault string of photovoltaic array based on current time series change detection

Abstract

Various faults inevitably occur in photovoltaic (PV) array due to the harsh external working environment. Therefore, detecting the faults and theirs locations is essential for the PV array. In this paper we propose a method for detecting the faults and their locations based on time series of PV string current. A time series sliding window (TSSW) is adopted. The local outlier factor (LOF) of each current point in the TSSW is calculated. Once a number of LOFs are continuously detected to exceed the threshold value, the PV string can be judged as fault. The experiment results show that the proposed method can detect short circuit fault, open circuit fault and shadow fault for PV string under different irradiance.