A Mixed Signal Separation Algorithm Based on Wavelet Transform
Xintao Jiao · 2018
When there is fault in the gearbox whose support is elastic, the vibration signal is complicated. Using the traditional methods for fault diagnosis may get wrong result. Based on the analysis of the characteristics of the steady modulation and impulse modulation signals, a mixed signal separation algorithm using the advantages of the inner product transform and the wavelet packet transform is proposed. Firstly the steady modulation components are extracted from the signal by taking inner production between the signal and the various basis functions. Then the impact modulation ones are gotten from the remaining parts of the signal by using wavelet transform. The simulation result and the experiment result show that the proposed method can extract the two kinds of signals effectively.