Research on Fault Diagnosis of Photovoltaic Array Based on Random Forest Algorithm

Abstract

Solar photovoltaic power generation has attracted widespread attention for its advantages of zero pollution, sustainability, flexibility, and high reliability. In order to ensure the normal operation of photovoltaic power plants and reduce serious accidents and power plant revenue losses caused by equipment failures. This article has carried out research on photovoltaic panel fault detection and diagnosis methods. This article first systematically summarizes the causes of photovoltaic array failures and summarizes the types of failures. In order to solve the problem of intelligent demand for fault monitoring in large-scale photovoltaic power plants, it is proposed to use the random forest algorithm in machine learning to establish a data mining decision tree model for photovoltaic panel operating data, and use the model to predict the cause of photovoltaic panel failure. First, collect the PV array current, output power, temperature sensor PV panel positive board working mix and other indication data in the PV array monitoring data. Further data extraction, transformation and loading of massive data are carried out to establish a photovoltaic array fault diagnosis application database. Finally, the random forest algorithm was successfully implemented, the model was established, the model was predicted, and the accuracy of the prediction result was high.