Fault Diagnose Of Rolling Bearing Research Based On Local-wave Decomposition Technology

In this thesis the fault diagnose of train bearing is studied theoretically and experimentally. Aiming at the limitation of local wave decomposition, the thesis proposes an improved arithmetic and verifies the correctness of it through digital simulation and experiments. The main content are summarized as follows:(1) To indicate the purpose and meaning of the study, the thesis summarizes the development of fault diagnose of train bearing. In second chapter introduces the fault types and structure of rolling bearings, then calculates the vibration shape and fault characteristic frequency, also introduces experiment platform and software.(2) In chapter three analyzes the basic principle and end effects of local wave decomposition, also introduces three methods to solve the end effects: extreme extension method, mirror extension method, endpoint matching feature wave continuation method. Using the method above to decompose the local wave by digital simulation. The simulation results show that the endpoint matching feature wave continuation method is better than the other two. The method abandons the traditional idea of removing the head and tail of a piece of data and combines the characteristic of the decomposed data to restrain the end effects. The method can also avoid the influence of distortion and make the characteristic frequency of component signal more obvious.(3) In chapter four analyzes the reason of mode mixing in detail and uses the improved EMD-ICA to avoid the inconvenience caused by exception event, which make the method of local wave decomposition more accurate and effective. Aiming at the false component of local wave decomposition, this chapter uses cross correlation coefficient and mutual information to simulate respectively. The results indicates that using mutual information can eliminate the interference of other components. The feasibility of this method can be proved by using it to analyze the fault signal of ring,roller and inner ring.(4) In chapter five designs fault diagnose of bogie bearing system. From basicconcept and characteristics to introduce the software platform of fault diagnose of bogie bearing system--virtual instrument. Then the chapter designs the precision mode based on the theory of improved local wave method. The system has a good man-machine interface and a strong ability of calculation by adopting the advanced MathScript node. Finally experimenting the bearings which contain inner ring, outer ring, roller failure as examples to verify the theory above.