Research on Topology Identification of Distribution Network Under the Background of Big Data

Abstract

Traditional identification of distribution network topology needs to rely on manual identification and verification, which has the problems of low identification rate and poor accuracy. Under the current background of power big data, effective results can be obtained by fully mining the potential effective information of big data and applying it to the identification of distribution network topology structure. This paper presents a line-to-line relationship identification method based on the maximum correlation minimum redundancy method and a line-to-transformer relationship identification method based on the conversion of three-phase unbalanced outlet voltage. The voltage acquisition system is fully utilized, and the maximum correlation minimum redundancy method (mRMR) is used to eliminate the redundancy of feature variables in line-to-line relationship identification to obtain the most accurate correlation results. In the identification of line-transformer relationship, the conversion method of three-phase unbalanced outlet voltage is adopted to eliminate the influence of misjudgment caused by three-phase unbalanced voltage and improve the identification accuracy. The identification method proposed in this paper is applied to the topology identification of a regional distribution network. By comparison with the actual topology, the identification accuracy is up to 99.78% and the effect is remarkable.