Anomaly electricity detection method based on entropy weight method and isolated forest algorithm

This study aims at investigating the applicability of abnormal electricity consumption data detection method, which is based on the entropy weight method and the isolated forest tree algorithm. The inaccessibility and imbalance of abnormal electricity consumption samples in actual data sets are considered by analyzing smart distribution network power consumption big data. Firstly, the users are classified by the k-means clustering algorithm, and then the characteristics of each type of user are extracted and the feature set is processed by the principal component analysis method to reduce the dimensions, followed by the entropy weight method adaptive configuration of the weight coefficients of each feature index, and finally the abnormal power consumption users are calculated based on the feature-weighted isolated forest algorithm. The algorithm verifies the real electricity consumption data of 6,445 users, and the results show that the method has a high detection accuracy, recall rate and F1 score, which is more suitable for the detection of abnormal electricity consumption in scenarios when there are complex and diverse user power consumption behaviors.