考虑分布式光伏的低压台区线损异常辨识方法

Line loss anomaly identification method for low-voltage station area considering distributed PV

低压配电网线损异常的辨识一直以来都非常困难,分布式光伏大量接入配电网,改变了配电网的潮流分布,更加大了低压配电网线损异常的辨识难度。提出了一种针对分布式光伏接入台区线损异常的辨识方法。首先对分布式光伏接入台区开展光伏出力等因素与线损率的灰色关联度计算,找寻光伏相关因素与线损率关联性。其次根据关联性的强弱选择合适的指标进行k-means聚类,并依据聚类结果进行离群点检测,判断台区是否有线损异常的可能性。最后通过对离群点所在簇进行时间离散度分析,得出台区的异常系数,根据异常系数进行线损异常判断。通过对含分布式光伏的典型台区进行验证分析,结果表明:该方法能够有效辨识分布式光伏接入台区的线损是否异常。

The identification of line loss anomalies in low-voltage distribution networks has always been very difficult,and the large number of distributed photovoltaics connected to distribution networks has changed their power flow.This makes the identification of line loss anomalies even more difficult.In this paper,a method is proposed to identify line loss anomalies in distributed PV access station area.First,the grey correlation between PV output and line loss rate is calculated to find the correlation between PV-related factors and line loss rate for distributed PV access station area.Second,k-means clustering is carried out to select suitable indicators according to the correlation of station area line loss,and outlier detection is carried out based on the clustering results to determine whether the station area has the possibility of a line loss abnormality.Finally,by analyzing the time dispersion of the clusters where the outliers are located,it can obtain the abnormal coefficient of the station area,and judge the abnormal line loss based on that coefficient.The results show that the method can effectively identify whether the line loss of a distributed PV access station is abnormal or not through the verification analysis of typical station areas containing distributed PV.