Abstract: Aiming at the problem of low accuracy and comprehensiveness of current identification methods，a

method of reverse power anomaly identification in digital electric energy meter information sampling was proposed.

Through the data acquisition system，the digital electric energy meter information sampling work was completed，and the missing value filling，data standardization and other pretreatment were implemented. Based on the processed data，the state features of the watt-hour meter were calculated，including the features of three dimensions，such as the change of the user's electricity consumption，voltage/current and active power，and AdaBoost algorithm was used to construct a classifier to realize reverse power anomaly recognition. The results show that under the application of the proposed method，the F1 score of the method is higher，the method can detect the electric stealing more accurately and comprehensively，which provides a reliable basis for the electric stealing user.

      Key words: digital electric energy meter；information sampling；reverse power；anomaly identification；features；AdaBoost algorithm