Time-Frequency Domain Analysis: Wavelet-Transform Based Fault Detection

A comprehensive analysis of wavelet-based method detecting and locating fault in HVDC system is detailed in this chapter. We study how wavelet transform can serve as a protection tool to decide right tripping signal. Detailed coefficient is one of the parameters we use to determine the fault occurrence. Using the analytical fault signal, we illustrate how the coefficient changes during pre- and post fault. There are many types of mother wavelets, for instance, Haar, Daubechies, Coiflet and Symlet. The choice of mother wavelet is a topic of interest here. Particularly, we find that Daubechies wavelet is the most optimum for fault detection as far as sensitiveness is concerned. There are some factors that can influence the performance of wavelet transform: fault type, fault resistance, fault location, choice of wavelets and sampling frequency. The simulation results show that the wavelet transform can tolerate these influences reasonably well. We use two types of simulation models: two-terminal and multi-terminal HVDC systems.