Intelligent Fault Classification Algorithm of Photovoltaic Array based on Noise Identification of Cluster Analysis

The development of solar photovoltaic power generation system has slowed down the aggravation of the energy crisis. Meanwhile, it has brought hidden trouble such as aging, short circuit, open circuit, hot spot of solar panels, which will cause serious obstacles to the power generation of solar photovoltaic system. Therefore, a timely and efficient online fault diagnosis method is the guarantee for the stable operation of photovoltaic system. In this paper, a 4x3 experimental model is built by a small photovoltaic power station to collect 180 sample data of each of the five types. Normalized preprocessing was performed on the data. Aiming at the situation that the experimental data classification accuracy is low due to the interference of human and environmental factors, a multi-fault classification and diagnosis optimization method of smart photovoltaic array based on cluster analysis noise identification is proposed. The optimized fault classification accuracy can reach more than 98%, which provides a favorable theoretical basis for photovoltaic array fault diagnosis.