Gear Vibration Signal Analysis Method Based On Wavelet Transform

To actual gear wheel vibration signal analysis, this dissertation has studied the vibration theory of the gear wheel at first, the project meaning to wavelet analysis theory, the application of wavelet transform to the signal on this basis. This text has finished the following several tasks mainly:First of all, through multianalysis and studying of the vibration theory of the gear wheel, draw the conclusion: effective time-frequency analysis is the inherent demand for analyzing the vibration signal of the gear wheel; singularity monitoring and filtering the signal or depressing noise are the main applications of wavelet to gear wheel vibration signal analysis.Secondly, through to study gear wheel vibration signal analysis course related to wavelet theory that arrives research, explain with various wavelet project meanings and theoretical foundation. Comparing wavelet analysis and other common signal analytical methods, and on the basis of comparing every performance index, choose Daubechies (N =1, 3, 6), Coifi (Coiflet wavelet system) wavelet as wavelet mother functions for gear wheel vibration signal.Third, prove the original advantage of wavelet transform through the signal instance that while dealing with the vibration signal of the gear wheel. And through gathering actual gear wheel vibration signal from laboratory, utilizing wavelet transform to study singularity monitoring, has made the good results.Fourth, directing against the characteristic of gear wheel vibration, have recommended drawing a certain signal with interval frequency in the signal, suppressed or decayed some signals with interval frequency in the signal, predicted the development trend of the signal, such contents as similar of analyzing the signal. Such also offer the effective analysis means for the thing that the vibration signal of the gear wheel will be analyzed and researched in the future.The fifth, through the analysis of the simulative signal and instance signal, expound wavelet analysis is going on advantage and characteristic on depressing noise.