Predictive Warning of Photovoltaic Generation Anomalies Based on Neighbour Profile Method with Spatial Correlation

Abstract

A photovoltaic generation has been widely applied due to the maturity of the correlation technique and the falling cost of materials. However, the characteristics of strong volatility and uncertainty have limited the consumption of photovoltaic power energy, which is caused by a combination of random factors from the natural environment and photovoltaic device conditions. Compared to the unpredictability of random natural factors, photovoltaic device anomalies can be detected from the spatial correlation. This paper proposed the photovoltaic generation anomalies warning method based on the neighbor profile method. To testify to the efficiency of the proposed method, experiments based on the four real photovoltaic plants are conducted in this paper. The results show the neighbor profile value of abnormal devices is different from the normal ones and verify the feasibility of the proposed method.