Study Of Single Phase-to-earth Fault Location Based On Wavelet And ANN For The Small Current Neutral Grounding System

The small current neutral grounding system is a kind of system without neutral grounding, in most of which the neutral point connects to ground through a reactor. When single-phase ground fault occurs in this system, the fault current is much low and the phase-phase voltage is still symmetrical. So the system still can supply power to load. Although many researchers have done much concerning on fault location for years, the key problem remain unsolved because of the undistinguished fault feature. With the requirement of improving the reliability of power supply and the fast development of the distributed power system automation, the study of fault location becomes more and more important.This paper analyses the steady and transient character of line fault when the neutral point connects to ground in different ways. From the angle of signal processing, the wavelet analysis theory and the artificial neural networks are introduced to fault location, in which the fault character is not the steady signal but the transient signal. Judging by the different magnitude and direction of the wavelet packet coefficient modular maximum, we can select the single phase-to-earth line. The method using wavelet singularity detection theory to extract the fault time is discussed. A fault distance measurement algorithm combining wavelet analysis used for separating the character from the transient fault signal with artificial neural networks used for the nonlinear approximation from the transient character to fault distance is presented. The data from a typical small current grounding neutral system example validate the feasibility of the algorithms.In spite of the limited samples, the simulation result shows that using Wavelet analysis to separate the character from the transient fault signals is an efficient way.