Photovoltaic Fault Diagnosis Research Based On CW-RNN Feature Analysis

In the context of double carbon,solar photovoltaic power generation,as one of the most important renewable energy sources,will continue to maintain rapid growth in scale.Because the photovoltaic power station is built in an outdoor environment,the panel itself is susceptible to environmental corrosion and is prone to multiple types of failures.In this paper,we study the photovoltaic fault diagnosis method based on CWRNN characteristic analysis,and build a fault analysis platform for photovoltaic power generation.The main tasks are as follows:Firstly,the photovoltaic cell engineering model is established by using the mechanism modeling method,and then a photovoltaic array simulation model including the photovoltaic array model,grid-connected control strategy and maximum power tracking control strategy(MPPT)is constructed in matlab environment;the causes of photovoltaic power generation failures are analyzed,and the output characteristic data of photovoltaic arrays under different fault conditions are simulated;at the same time,based on the photovoltaic array measurement and control experimental system,the output characteristics of photovoltaic arrays under different fault types are simulated.Provides a data source for subsequent studies of troubleshooting issues.Secondly,based on the output data of photovoltaic panels of different fault types,the parameters of the fault characteristics of photovoltaic arrays are obtained.After normalizing the dataset as the input of the algorithm sequence,the recurrent neural network(RNN)algorithm and the clock recurrent neural network(CW-RNN)algorithm are used to realize the forward propagation of the algorithm process and the cyclic feedback model based on time backpropagation,so as to realize the identification and localization of the fault of the photovoltaic array,and further elaborate the fault diagnosis process of the two types of neural network algorithms.Finally,a fault analysis platform for photovoltaic power generation including the bottom layer,the middle layer and the upper layer is built.The bottom layer includes the Simulink simulation model and the photovoltaic array measurement and control experimental system;the middle layer includes the real-time database system and the calculation engine;and the upper layer is the visual monitoring interface.Based on the platform,the characteristics and parameters of fault diagnosis are displayed,and the simulation results are analyzed to verify the effectiveness and accuracy of the CW-RNN algorithm model for fault diagnosis.