FAULT SIGNALS DENOSING AND COMPRESSING BASED ON MBDW FILTERS
Ren Zhen · Proceedings of the CSEE · 2001
As the levels of WT increasing, WT coefficients of noises are trending toward zero rapidly,so MBDWs can be used to denose in this paper. Firstly, according to band width define,the MBDW filter co efficients of L=4,N=6,8 were calculated and tabulated. Then the stator fa ult current signals of the motor are denosed and reconstructed, meanwhile, the b roken-bar fault current signals are compressed and reconstruced. Three indexes of the reconstructed signals, including relative error, energy recovering coeffi cient and compressing rate, show that MBDW-based signal processing method is ef fective,low complex,rapid and real time.