基于信息熵的自适应窗长光伏系统直流串联电弧故障识别研究

采用对电流参数最终归一化的时域分析和基于信息熵自适应选取窗长的频域分析相结合的方法，研究了光伏发电系统直流串联电弧故障的识别问题。通过设计自适应窗函数，获得直流侧电流归一化的时域信息和基于信息熵自适应选取窗长的频域信息。根据串联电弧故障状态和正常工作状态下的特征差异，提出了用时频域信息结合以提高识别灵敏度的方法，经实验验证，该方法能可靠地识别各种状态下的串联电弧故障，不受阴影遮挡等条件的影响，可在保障光伏发电系统安全运行中发挥积极作用。

Study on DC Series Arc Faults Identification for Photovoltaic System with Adaptive Window Length Based on Information Entropy

In this paper, the identification of DC series arc faults in a photovoltaic(PV) power generation system is studied by combining the time-domain analysis of the final normalization of current parameters with the frequency-domain analysis of window length adaptive selection based on information entropy. Through the design of the adaptive window function, the time-domain information of the normalized DC current and the frequency-domain information of the window length adaptive selection based on information entropy are obtained. According to the characteristic difference between the series arc fault and normal working states, the method of combining the time- and frequency-domain information is proposed to improve the identification sensitivity. It was experimentally verified that this method can reliably identify the series arc faults under different conditions and it was not influenced by conditions such as shadow, which can play a positive role in ensuring the safe operation of the PV power generation system.