Fault Detection and Location for VSC- HVDC Systems with the Coexistence of AC and DC Faults by Using PSO- Neural Network

Abstract

Superior power quality and less transmission loss are features of HVDC transmission. However, fault location and clearing technology are less advanced compared to their AC counterparts. The convergence rate and classification accuracy of the learning process can be enhanced by particle swarm optimization (PSO); the PSO method has been selected and employed in this study's feedforward neural network. This paper proposes a fault localization technique for a 75-kilometer, two-terminal VSC-HVDC system by using the wavelet transform (WT) and the PSO with artificial neural networks (PSO- NN). The HVDC system is simulated with MATLAB, and the results are then processed with WT, PSO-ANN, and both traditional and conventional ANN. The eight DC faults in various places and the simultaneous AC faults are the faults in this study. To examine the impact of these two factors on the suggested fault location approach, the simulation is run while varying fault resistance and location along a 75-kilometer distance. The result shows that it is possible to anticipate the fault location pretty reliably and with little error. Additionally, PSO-ANN performance has demonstrated its advantage over conventional ANN.