Application of improved WMRA in diagnosis of rolling bearing defects
                                                                      LIN Si miao, ZHANG Yan rong, GUO Li ping
                                   (School of Mechanical Engineering, Southwest Jiaotong University, Chengdu 610031, China)


Abstract: Aiming at the problem of the weak signal of defects and the traditional envelope spectrum needed to be estimated its frequency band through experience, the combination of empirical mode decomposition and optimized WMRA were put forward. Firstly the vibration signal was resolved through EMD, and the kurtosis and the characteristic parameters of vibration natural frequency were used to classify the component of intrinsic mode function，to screen out its the optimal component whose envelope signal can be calculated through Hilbert transform. The wavelet multi resolution analysis was used to reconstruct the envelope signal to obtain the envelope spectrum which presented the frequency of characteristics of defects obviously.The result of the research indicates that the method proposed can be used to do visualization analysis of time and frequency domain of signal of defects more effectively, when compared with the solo EMD or WMRA.
Key words: bearing defects; empirical mode decomposition(EMD); wavelet multi resolution analysis; intrinsic mode function; Hilbert transform