Anomalies in photovoltaic module(PV module)will directly affect the power generation and the service life of PV module.Identifying and eliminating anomalies in PV module timely will directly improve the power generation efficiency of PV module.In order to accurately detect the abnormal conditions of PV module,6 types of abnormal conditions of PV module are simulated based on Matlab platform and the output characteristics are analyzed,and the five-feature method for output characteristic curves are proposed to judge the type of abnormal conditions.Python language is used to establish the probabilistic neural network and particle swarm optimization algorithm is adopted to optimize the smooth factor.The particle swarm optimization-probabilistic neural network(PSO-PNN)model is used to train abnormal data.The result shows that,the probabilistic neural network is sensitive to the dataset.For a large dataset,the model classification has a high accuracy,which can effectively detect abnormal conditions of PV module.
维普
万方数据