Distribution Network Reconfiguration Based on Relevance Vector Machine

Abstract

The distribution network reconfiguration can be described as a pattern recognition problem, a method to solve this problem, which is based on the strategy of building relevance vector machine pattern recognition model of distribution network is proposed. Having better generalization ability and faster training speed, the model takes the load pattern as its input and outputs the state of the branch group after the distribution network is simplified. Finding the exact location of the branch group with connective switches by using the proposed model, then according to the improved optimal flow method to find the position of the connective switches, so as to find the optimal structure of distribution network with minimum active power loss. Results show that the proposed method can obtain a more accurate recognition result in small and medium scale distribution network.