Nondestructive Testing Of Gearbox Based On Wavelet Analysis

Modern mechanical equipment increasingly high demands on performance, its structure is more and more complex,and the large-scale mechanical equipment and automation is increasingly developing.Gearbox,as a common parts using in transmission power and connection in machinery and equipment,is widely used in modern industrial equipment. In engineering practice,we expect that we have a preventive measure for the occurrence of gearbox failure,we don’t want to deal with incidents after the occurrence of gearbox failure has happened.It has important social significance and economic value to study effective gearbox fault diagnosis technology and quantitative diagnosis of degree of fault and predictive failure extended trend and remaining life of gear box.Wavelet analysis is currently a rapid development of new areas in mathematics,it has been widely used in many engineering,becoming one of the tools that scientists often use.This paper studies the work for theoretical research and analysis around the application of wavelet analysis in fault diagnosis of gearbox.The focus of this paper is to determine the gear crack form of this failure which is the most difficult to identify and has great potential threat.First of all,this paper studies the failure mechanism of gearbox and the basic principles of wavelet analysis,then studies the work to make a comparison between wavelet analysis and other signal analysis method,and makes a introduction on application of wavelet analysis in filtering,denoising and detection of mutation signal. Finally by example,this paper works on the analysis of gearbox fault signal through the method of wavelet analysis on Matlab sofeware.This paper uses db5wavelet function for the analysis of collected vibration signals and uses threshold denoising to extract the single frequency signal, then judges that the original signal is a fault signal and points out the failure form is gear crack. Then using the Fourier analysis to process collected vibration signals,this test indicates that the deficiencies of the Fourier analysis in dealing with non-stationary signals.Finally this article uses the bior5.5wavelet analysis to process collected vibration signals and suggests that the db5wavelet processing is better than bior5.5wavelet by comparing the proposed use of wavelet analysis in signal processing, and puts forward the principles of selection of the wavelet function when using wavelet analysis to deal with non-stationary signals.Wavelet analysis,as a new discipline which developed rapidly in recent years,has the double meaning both in deep rich theory and in-depth application.Its application and the theory development closely intertwined and mutually reinforced.In this paper,wavelet analysis theory is applied to direction of gearbox fault detection and achieves good results.It plays an important role in ensuring the safty and reliable operation of the gearbox equipment.Wavelet analysis has broad application prospects and this new technology is a worthy of popularization and application.