The power grid in China has been developed into a combination of information network, computer network and power network. The existing power information detection technology is mainly aimed at power generation, transmission and substation, and lack of effective power information detection means in the distribution terminal. Therefore, this paper proposes an analysis method based on PLR time series data extraction and VOD anomaly data detection. In this scheme, PLR algorithm is used to realize the linear segmentation of the original power data sequence and obtain the new data sequence. Then used the VOD to partition the new data sequence in two dimensions, and judge whether there is data Outlier Based on the outliers function generated by the point set. The simulation results show that the scheme can effectively identify outliers in the data sequence and provide auxiliary decision-making basis for the power workers.