Fault Location on Radial Distribution Systems Using Wavelets and Artificial Neural Networks with a New Data Processing Feature

Abstract

Every power system is vulnerable to fault occurrences. Under permanent fault conditions, maintenance crew has the duty to detect the problem, repair the defect and recover the power supply system. If the fault is previously located, the repair can be performed faster. The common methods to locate the fault in electrical distribution systems use the final user information or some heuristics with fuse coordination and the loss of loads. In this paper, a new algorithm for fault location in distribution power systems is presented. Using computational simulations, travelling waves theory, wavelet transform, a new data preprocessing feature, and artificial neural networks, this new algorithm tries to approximate the fault location using data provided by only one measurement point at the beginning of the feeder.