Abstract: With the continuous advancement of China's energy Internet strategy，how to reduce the occurrence

of electricity theft has become a focus of research. The historical data of electricity consumption was processed and

analyzed，and the data of electricity consumption were processed and converted to realize the integration of the data of electricity consumption. Then，through the anticipating analysis of users' electricity usage data，the key

characteristic indicators of electricity theft behavior were excavated and a identification method of electricity theft

users based on GA-BP neural network was proposed. Finally，the real-time operation data of power system and the

user identification method were used to realize the identification and diagnosis of electricity theft behavior. The

rationality and effectiveness of the method were verified by analyzing the actual operation data of the power system，which improved the stable operation of the transmission line and ensured the safe operation of the energy Internet.

      Key words: energy Internet；electicity theft；data mining；genetic algorithm-back propagation（GA-BP）neural network