The Signal Analysis Of Roller Bearing Fault Diagnosis Based On Wavelet Packet And EMD

The roller bearing is not only one of the most ordinary in the rotating machinery, but also is liable to be damaged. It's of great realistic significance to study fault diagnosis technical for it. In the process of fault diagnosis, the key point is extracting fault feature from the fault vibration signals. And the modern signal processing technology used in the extracting process plays a central role.The traditional time-frequency analysis method has a variety of limitations for the non-stationary and non-linear signals as the roller bearing vibration signals. In this thesis, the characteristics and limitations of the time-frequency analysis such as the short time fourier transform, Wigner-Ville distribution and wavelet transform methods are analyzed. Followed by the introduction of the instantaneous frequency, the intrinsic mode function, the Hilbert spectrum, and other basic concepts, it's described the empirical mode decomposition method (EMD) and the principle of algorithm created by the Chinese American N. E. Huang and others in detail, and revealed that EMD decompose signal into a series of single-component elements. The Hilbert transform as effective bridges reflected signal frequency, amplitude changes in the local laws of the instantaneous nature. And it's tested the Adaptability, multi-resolution, the completeness of EMD.According to the problems such as mode mixing and undesirable intrinsic mode functions (IMFs) at low frequency region EMD generated, it's proposed two types of outer and inside combination methods of wavelet and EMD to improve the sifting of EMD. The outer method using the wavelet denoising advatnage and combniationg with EMD based on wavelet denoising solves the problem of noise interference. The inside method starting with definition of intrinsic mode function, introduces a sifting method based on wavelet packet. It avoids a series of problem caused by applying cubic spline interpolation to sift IMFs from signal, and increases the adaptability and efficiency of signal decomposition. After testing by simulated signals and the rolling bearing fault vibration signal, the feasible and effective of the method proved.