Application Research of Artificial Intelligence Technology in Error Diagnosis of Electric Energy Meter

In China's power industry, the power meter error diagnosis is generally carried out through manual verification, operation sampling inspection, user application and other manual methods. However, with the popularization of automatic collection of power data, the amount of data is getting larger and larger, and the data is getting more and more ample. It is possible to realize error recognition through data analysis and artificial intelligence technology base on big data. The pass rate of the three sampling companies in Shanghai, Chongqing and Tianjin all exceeds 99.99%, and the State Grid Shanghai Electric Power Company users apply. The school pass rate is 99.16%. State Grid Shanghai Electric Power Company tested the long-running re-entry electric energy meter. The test data showed that the electric energy meter with the measurement error within 60% error limit accounted for 96.35%, and the electric energy meter with the measurement error within the qualified range accounted for 98.33%. 1.67% of the errors were unqualified. The statistical results show that the error diagnosis by manual method is inefficient, and a large number of operating energy meters are qualified, which wastes human and material resources and can’t locate the fault table in time. Therefore, it is necessary to apply big data analysis and artificial intelligence technology to study the data mining and calculation of the electricity consumption of the users in the Taiwan area, the total meter power of the Taiwan area, the relationship between the households in the Taiwan area, the line loss, and the user files. The theoretical model of error analysis of intelligent electric energy meter improves the accuracy of on-site fault judgment algorithm, and carries out large-scale pilot application and effect analysis.