Fault location in distribution network with distributed generation based on neural network

摘要

In this paper, the fault location problem of the distribution network with distributed generations(DGs) is studied. A fault identification and location method based on multi-stage neural network is proposed. Particle Swarm Optimization (PSO) algorithm is employed to optimize the network structure. The model and algorithm realize the fault location accurately. And when the input variables increases, When the input variables increases, its size is much reduced through gradual decomposition of the network compared to traditional the single-stage network. Finally, a typical distribution network with distributed generation is established in the Digsilent environment. The sample data are obtained through simulation experiment. The neural network model is established in Matlab to get the algorithm results. The experiment verified the correctness and effectiveness of the model and algorithm.