Fault Location in Transmission Lines Using BP Neural Network Trained with PSO Algorithm

摘要

In modern protection relays, the accurate and fast fault location is an essential task for transmission line protection from the point of service restoration and reliability. The applications of neural networks based fault location techniques to transmission line are available in many papers. However, almost all the studies have so far employed the FNN (feed-forward neural network) trained with back-propagation algorithm (BPNN) which has a better structure and been widely used. But there are still many drawbacks if we simply use feed-forward neural network, such as slow training rate, easy to trap into local minimum point, and bad ability on global search. In this paper, feed-forward neural network trained by PSO (particle swarm optimization) algorithm is proposed for fault location scheme in 500 kV transmission system with distributed parameters presentation. The purpose is to simulate distance protection relay. The algorithm acts as classifier which requires phasor measurements data from one end of the transmission line and DFT (discrete Fourier transform). Extensive simulation studies carried out using MATLAB show that the proposed scheme has the ability to give a good estimation of fault location under various fault conditions.