Selection strategy among categories of evaluation indexes for distribution network operation level based on mutual information

Aiming at the problems of many types of grid operation indicators, complex secondary indicators, and difficulty in qualitative or quantitative analysis, this paper proposes a grid operation indicator screening method based on mutual information with maximum correlation and minimum redundancy coefficient. First, sort out the relevant indicators of power grid operation evaluation at home and abroad, and then qualitatively screen the secondary indicators under the indicator category from the aspects of system operating efficiency, reliability, and economy. Secondly, based on the mutual information theory, the group correlation entropy values of different secondary indicators are obtained, and the maximum correlation-minimum redundancy coefficient is used to screen the entropy values of related index categories with the optimal correlation degree. These entropy values can maximize the mining of different The relevance of the secondary indicators under the category can minimize the redundant information between. Finally, it is verified by scoring a typical regional power grid operation. The results show that the proposed index correlation screening method can simply and effectively evaluate the operation level of the smart grid.