Jamal Moshtagh - Kurdistan University
Parham Jalili - Kurdistan University

The electric power markets have imposed new power service quality that makes fault detection in power distribution systems a mandatory issue. This paper presents a new approach to discriminate HIFs from transients such as load switching (high/low voltage) and inrush current, based on a new modified cable model, in the EMTP software. The simulated data is then analyzed using advanced signal processing technique based on wavelet analysis to extract useful information from signals and this is then applied to the artificial neural networks (ANNs) for detecting arcing faults in a practical underground distribution system. The paper concludes by comprehensively evaluation the performance of the technique developed in the case of arcing faults. The results indicate that the fault detection technique has very high accuracy

Fault detection, arcing faults, underground distribution cable, wavelet, neural network