Accurate fault location in high voltage transmission systems comprising an improved thyristor controlled series capacitor model using wavelet transforms and neural network

Abstract

This paper describes the applications of discrete wavelet transforms (DWT) coupled with conventional artificial neural networks (ANN) to the development of a fault location technique under an improved TCSC transmission system model. The fault location scheme is modular based whereby fault type is verified before identifying the fault location using ANN. This method relies on utilising DWT to decompose the busbar voltages and line currents obtained from a single terminal into a series of time-scale representations. A feature model using self-organising maps (SOM) is applied herein to verify the fault location capability of the extracted features. Results of computer experiments with simulated faulted TCSC transmission line are included and they indicate that this approach can be used as an effective tool for accurate fault location in TCSC systems. More importantly, it obviates knowledge of firing angle of the TCSC or any predefined assumptions.