Selection Method of Smart Grid Operation Index Based on Maximum Relevance and Minimum Redundancy

Abstract

Smart grid operation data presents the characteristics of large amount of data, multiple types of data and rapid growth rate, which is not conducive to quantitative analysis of actual operation level of power grid. In order to quickly and effectively screen the smart grid operation indicators, simplify the operation index system and reduce the index redundancy, this paper proposes a quantitative analysis index to measure the effectiveness and redundancy of operation indicators based on the maximum relevance minimum redundancy criterion. Then, through quantitative analysis of indicators, the optimization model is constructed. Finally, the screening process is transformed into a 0–1 programming problem by coding, and the optimal index set is obtained by exhaustive method. The results show that the proposed screening method improves the information content of the index system and reduces the index redundancy, which verifies the rationality and effectiveness of the screening method.