基于人工蜂群算法优化RBF神经网络的光伏组件故障诊断定位方法

Optimizing RBF Neural Network Fault Diagnosis and Location Method for PV Module Based on ABC Algorithm

随着光伏装机容量的迅速增长以及国家光伏补贴的逐步取消,针对提高光伏发电效率与收益的研究具有重要意义。首先,为精确定位光伏故障以降低运维成本,提出以组串支路电流与组件温度为故障定位特征量。其次,为提高光伏故障诊断效率,利用人工蜂群算法(artificialbeecolony,ABC)优化RBF神经网络的中心、函数宽度及权值,提出将其应用于光伏组件故障诊断定位研究。最后,根据盐城市响水县1.5MW光伏电站数据,经Matlab仿真研究并与传统RBF神经网络相比较,结果验证了所提理论的准确性。

With the rapid growth of photovoltaic installed capacity and the gradual elimination of national photovoltaic subsidies, it is of great significance to study how to improve the efficiency and benefits of photovoltaic power generation.Firstly,in order to locate photovoltaic faults accurately and reduce the cost of operation and maintenance,a series branch current I and component temperature T are proposed as fault location features.Secondly,in order to improve the efficiency of photovoltaic fault diagnosis,this paper uses artificial bee colony (ABC) algorithm to optimize the RBF neural network center,function width and weight,and proposes to apply it to the research of photovoltaic module fault diagnosis and location.Finally,according to the data of 1.5MW photovoltaic power station in Xiangshui County,Yancheng City,and the accuracy of the proposed theory is verified by the simulation study of MATLAB and comparison with the traditional RBF neural network.