A fault detection and diagnosis technique for solar system based on Elman neural network

Facing the shortage of energy supply and environmental pollution, the solar energy industry is developing rapidly all over the world. However, the harsh and volatile environmental factors will lead to many fault types in PV systems. Simple monitoring and fault diagnosis technology can not realize the requirement of intellectualization and informatization of PV power generation systems. In this work, we present a new detection and fault diagnosis method based on Elman Neural Network (ENN). The proposed method mines the implicit mapping relationship between original data and fault types and establish multiple hypothesis models, then compares the mean and variance of diagnostic errors to choose best diagnostic model. Based on the self - made photovoltaic power plant simulation platform to simulate the failure situation, the effectiveness of the new method was verified. The experimental results show that the proposed method can identify fault states in changing environment. Therefore, the new method to achieve the intelligent monitoring of photovoltaic power plant system to reduce the risk of power plant failure to improve power generation efficiency, with the promotion and application of significance.