Fault Diagnosis Of Photovoltaic Modules Based On Radial Basis Function Neural Network

The large consumption of traditional fossil energy caused serious environmental pollution and threatened the development of human society and ecological environment,so people begin to pay attention to the development and application of new energy sources.Among them,solar energy has been rapidly developed and widely used because it is clean,pollution-free,easy to install,geographically unlimited,safe and reliable,and can be used forever.As the most important part of photovoltaic power generation system,the service life and safety of photovoltaic modules have attracted more and more attention.If the faulty photovoltaic array is not diagnosed and handled in time,it may cause serious consequences such as fire.Therefore,it is of great significance to study the fault diagnosis of photovoltaic modules.In this paper,the photovoltaic module is taken as the research object,and the possible faults during the operation of the photovoltaic module are studied.Based on the Radial Basis Function(RBF)Neural Network's ability of pattern recognition,RBF Neural Network is selected to construct the fault diagnosis model in this paper.Particle Swarm Optimization(PSO)is used to optimize the center,width and connection weights of the hidden layer basis function of RBF Neural Network,which can avoid the problem of improper selection of parameters and better construct RBF Neural Network.The main contents of this paper are as follows:(1)Major faults of photovoltaic components,such as short circuit faults,open circuit faults,shadow faults and abnormal aging,are analyzed,and the causes of these faults are summarized.(2)Based on the simulation model of photovoltaic cell in MATLAB,the simulation model of 2×20 photovoltaic array is constructed,and different fault states are simulated by the simulation model.(3)Starting from the parameters of the PSO algorithm,the dynamic adaptive inertia weights and the adjustment learning factor and the update formula are used to optimize the parameters of the neural network.The RBF neural network model optimized by the improved PSO algorithm is applied to the photovoltaic array fault diagnosis process and compared with the RBF neural network and the basic PSO-RBF neural network model.The simulation results show that the fault diagnosis effect of IPSO-RBF neural network model is better than the other two models.(4)The parameters of RBF neural network are optimized by combining PSO algorithm with artificial fish swarm algorithm(AFSA).The AFSA-PSO-RBF model is applied to the photovoltaic array fault diagnosis process.The simulation results show that the diagnostic accuracy is higher than that of IPSO-RBF model.