Back propagation neural network aided wavelet transform for high impedance fault detection and faulty phase selection

High impedance fault (HIF) is very common problem and complex phenomena, and because of its distinctive characteristic is considered as riskiness for public safety and human. Therefore, the detection and protection of such faults still remain a topic of research and challenging of protection engineers. In this paper, a new model of (HIF) is introduced and tested with applying of new hybrid algorithm using the wavelet transform and the neural network. The Discrete wavelet transform (DWT) is used as feature extraction to extracts useful information from the distorted current signal that is generated from transmission system network under effect of the simulated model of (HIF). In order to improve training convergence and to reduce the number of inputs to the neural network, the coefficients of wavelet are calculated and used as the inputs for training Multi-layer back propagation neural network (BP-NN) for detection the high impedance fault and discriminate the faulty phase from healthy one.