Research on Vibration Signal Filtering Based on Wavelet Multi-resolution Analysis

It is very important to enhance the real-timeliness of vibration active control system, and the efficiency and effectiveness of vibration signal filtering is the one of the important factors to the real-timeliness. Multi-resolution wavelet analysis of vibration signals filtering is appropriate. Based on the theory of wavelet multi-resolution analysis, adopted Mallat signal decomposition and reconstruction algorithm, selected Db6 wavelet function, multi-resolution decomposed the vibration signal with scale 5, processed the high frequency signal according to the compliably filter method, reconstructed the vibration signal, and successfully realized to the vibration signal filtering. Emulated with the vibration signal of rotor experiment platform as an example, filtering processed the signal, and the effect of signal filter was very obviously. It is shown that the method of wavelet multi-resolution analysis is very suitable for the analysis of vibration signal, reduced the calculation work, enhanced the calculation efficiency, and easy to the vibration active control real-time process.