Abstract: This article starts from the centralized verification of the dismantled electric energy meters at the provincial level. Based on the data of the first inspection, operation, and disassembly of the full life stage of the intelligent electric energy meters, the power big data analysis application platform is used for data processing, modeling, and analysis. The research focuses on the aspects of automatic sorting of the dismantled electric energy meter, multi-dimensional analysis of fault characteristics of the energy meter and fault prediction of the electric energy meters, and the results can be achieved by graphical display. The application of related achievements can predict the occurrence of energy meter failure to a certain extent in advance, and guide the accurate replacement of the energy meter in operation, which can increase economic and social benefits.