Key node identification and network simplification modelling method for optimal power flow analysis of active distribution network

Abstract

With the increasing scale of active distribution network (ADN), the power flow analysis of ADN faces with many difficulties, such as insufficient online measurement and high complexity of overall modelling. Therefore, this paper proposes a key node identification and network simplification modelling method for the optimal power flow (OPF) analysis of ADN. Firstly, a voltage extremum similarity index is proposed and combined with K-means clustering algorithm to partition the ADN into several clusters, and then a comprehensive evaluation index is constructed with the entropy weight method to identify the key nodes of each cluster. On this basis, a key node reduction procedure is established with the comprehensive index of voltage violation probability to further reduce the scale of key nodes. After that, the power supply paths between key nodes are searched by the depth-first search algorithm to construct the topology of simplified network and then an improved π-type simplification network method is proposed with load displacement principle to establish the simplified network model with low dimension. Finally, the modified IEEE 123-bus system is used to verify the effectiveness and accuracy of proposed method. The simulation results indicate that the proposed key node identification method can accurately identify the voltage extreme nodes including PV connected nodes, and the proposed network simplification method can effectively improve the efficiency of OPF analysis with guaranteed accuracy.