Abnormal line loss identification of distributed PV low-voltage distribution area based on data driven

Abstract

The identification of line loss anomalies in low-voltage distribution networks has been a very challenging problem for a long time. The large-scale access of distributed photovoltaics to the distribution network changes the power flow distribution and further increases the difficulty of identifying abnormal line losses in low-voltage distribution areas. In this paper, an identification method for abnormal line loss in distributed photovoltaic access low-voltage distribution area is proposed. Firstly, we analyze the correlation between the line loss rate and other indicators in the area of distributed photovoltaic access. Secondly, according to the gray correlation degree results, the appropriate indicators are selected, and the k-means clustering algorithm is used to cluster these indicators, through the clustering results, we can detect outliers to determine whether there is a line loss anomaly in the area. Finally, by analyzing the time dispersion degree of the cluster where the outlier point is located, the anomaly coefficient of the area is obtained, and the line loss anomaly is judged according to the anomaly coefficient. To verify the effectiveness of this method, we conduct an empirical analysis of typical containing distributed photovoltaics. Experimental results show that this method can effectively identify the line loss abnormalities in the distribution area containing distributed photovoltaics.