Data Desensitization Method of Electricity Information

Abstract

With the explosive growth of electricity information, its potential value and security problem is becoming more and more attractive. To ensure the safety and reliability of electricity data at different application levels, this paper proposes two kinds of data desensitization methods based on Fourier transform theory by studying the appropriate processing measurements, connotation and adjustable parameters. Specifically, recoverable and irreversible desensitization methods are proposed, which are used in the typical scene of electric data mining—load clustering. Experiments show that both types of methods can support clustering analysis effectively.