Research On Application Of Fractal And Wavelet Analysis In Fault Diagnosis Of Hydraulic Pump

Hydraulic system has been extensively used by its excellence such as high power and quick response in engineering area and it plays an essential role in control and transmission of many equipments. The pump is the "heart" of the hydraulic system, and it has the severest working condition. So the condition monitoring and fault diagnosis of the pump is crucial for the hydraulic system. In recent years, the chaotic engineering is becoming the subject being paid attention abroad and the research on chaos and fractal can help to find out the internal regularity of system. In this paper, the fractal theory is applied in fault diagnosis of the pump and the correlation dimension of pump pressure signal is calculated. The results of test show that there is different correlation dimension in different working state, which can be a proof of whether fault occurs or not.The main works in this dissertation are showed as follows.(1) The feature of chaos is described in detail. New development of chaos and fractal in fault diagnosis field is summarized. The algorithm of correlation dimension (G-P algorithm) is expatiated and how to choose the parameter of it is given.(2) The nature character of wavelet analysis is expatiated. Wavelet analysis method is used in the process of state signal of hydraulic pump. Through the multi-resolution decomposition, threshold value elimination and reconstruction, the high frequency noise in the system is excluded and the characteristic frequency of fault signal is abstracted. (3) The feathers of virtual instrument and its development tool LabVIEW are summarized systematically. The interface of LabVIEW and MATLAB is programmed. This mixed programming extends the function of LabVIEW. The virtual chaos character shower is developed through which the chaos character is analyzed. The moving morphologic identifier based on the method of correlated plane is developed through which the moving morphologic of chaos is identified with others.(4) LabVIEW is used to design the data acquisition software of hydraulic pump. With the soft the NI DAQ card is driven and the signal of hydraulic pump is measured.(5) The obtained pressure signal of pump outlet is analyzed with fractal dimension. The calculated results show that the correlation dimension of fault signal is lower than that of normal signal. This conclusion can be a good quantitative evidence of whether fault occurs or not.