| With the complexity , automation and continuous improvement of equipment, the failure and fault of rolling bearing and gear will result in the increasing loss of whole production,and even more likely to endanger the safety of the operator’s life. Therefore, it is very important and urgent to study and explore a new method of fault diagnosis for rotating machinery.In this paper, through the analysis the fault signal of bearing and gear in time domain we can find that the kurtosis factor, the peak factor, the margin factor, the waveform factor, and the pulse factor are sensitive to the fault signal. So these parameters can be used as the time domain characteristic parameters of rolling bearings and gears. By means of Fourier transform and auto power spectrum analysis, the natural frequency and the frequency band range of the fault signal can be determined. Time frequency analysis is an effective method to deal with non-stationary signals. Therefore, this paper studies the method of resonance demodulation based on wavelet packet.Through the wavelet packet decomposition of the fault signal, the low frequency wavelet packet coefficients are selected to reconstruct and have resonance demodulation analysis of reconstructed signal. The experimental results show that the method can find the characteristic frequency more accurately. At the same time, the BP neural network is used for fault diagnosis, and the fault features of the fault signal are used as the input of neural network. Experiments show that the method has better recognition effect.In this paper, the bearing and gear fault diagnosis system is developed based on the LabVIEW platform. The system includes signal playback module, time domain feature extraction module, frequency domain feature analysis module, resonance demodulation feature extraction module based on wavelet packet and fault diagnosis system module based on BP neural network. The system can identify and diagnose the fault signal, and has practicability and portability. The accuracy and stability of the fault diagnosis system are verified by experiments. |
