The Study of Fault Signal Detection Based on Wavelet Energy Entropy

摘要

In view that the power fault signal is catastrophic and non-stationary, how to detect the location for the catastrophe point of the fault signal is researched, the fault signal is pre-processed using a de-noising threshold method in this paper, and the wavelet and entropy is combined, and is applied to power system fault detection and location, and it provides a new idea and method for a singular signal. Through simulation analysis, some conclusions were derived as follows: the catastrophe point of fault signal is well detected through Wavelet Energy Entropy WEE ;if only the left side of the data window is in coincidence with the signal catastrophe point, WEE curve started to flatten, and the moment corresponded to the signal catastrophe point.