基于概率神经网络的光伏阵列故障诊断

Research on Fault Diagnosis of PV Array Based on PNN

针对光伏阵列发生故障时难以诊断的问题,运用概率神经网络对光伏阵列进行故障诊断。首先,运用MATLAB中的simulink模块建立光伏阵列仿真模型;然后,针对光伏阵列在开路、短路、局部阴影和异常老化等故障下的输出特性进行分析,从而每个电气参数在各种故障状态下的特征信息被提取;最后,通过建立概率神经网络模型对光伏阵列进行故障诊断分析。仿真结果表明,概率神经网络在光伏阵列故障诊断方面具有较好的准确性、适应性和有效性,为光伏阵列实际故障诊断提供了一种理论依据。

Aiming at the problem that it is difficult to diagnose the failure of photovoltaic(PV)array,probabilistic neural network is used to diagnose photovoltaic array fault.The simulation model of PV array is build by using the simulink module in MATLAB.Then,the output characteristics of PV array under open circuit,short circuit,partial shadow and abnormal aging are analyzed,and the characteristic information of each electrical parameter under various fault conditions is extracted.Finally,a PNN model is established to analyze the fault diagnosis of PV array.Simulation results show that PNN has good accuracy,adaptability and effectiveness in fault diagnosis of PV array,which provides a theoretical basis for practical fault diagnosis of PV array.