Sensorless Operation of a PMBLDCM Drive Based on Daubechies wavelet Decomposition Technique

摘要

Wavelet analysis is used in various application areas of electrical machines, due to its inherently suited for non-stationary signal analysis. Discrete forms of wavelet transform are necessary in most computer implementation; it is a signal-processing tool that decomposes the signal into several other signals of different scales with different time and frequency resolutions. Due to multi resolution analysis (MRA) of DWT, it is considered more suitable to the study of the signals, which have low frequency components with fast changes. In this paper, a multi-resolution technique using Daubechies wavelet decomposition method of the stator current signal of a permanent magnet brushless DC motor drive (PMBLDCM) drive is proposed. This method can be used successfully to predict the commutation instants of a PMBLDCM drive without any position sensor. Digital simulations based on Matlab/Simulink are implemented on the phase current and recorded in time domain. These waveforms were filtered by wavelet decomposition method using six levels of db6 (Daubechies). The commutation instants have been successfully captured without any position commutation sensors. It should be emphasized that the wavelet analysis is considered as a good tool for sensorless PMBLDCM drive.