Research On The Fault Diagnosis Methods Of Gears Based On Resonance-based Sparse Signal Decomposition

Gears are the key component for transmitting power and rotating motion in mechanical equipments. They are widely used in the mechanical processing machines, electric power systems, mining machineries and other modern industrial equipments. The fault of gears and gear boxes affects the safety and reliability of an equipment directly. Therefore, condition monitoring and fault diagnosis of gears has important significance and value.An important means for gear fault diagnosis is to extract the fault feature information from the fault gear vibration signals. Resonance-based sparse signal decomposition method is a new kind of vibration signal analysis method put forward by Selesnick. In resonance-based sparse signal decomposition method,the quality factor Q is defined as the ratio of the center frequency and bandwidth. According to that the quality factor of a harmonic component is different from that of a transient impact component, the method can decompose a signal into high resonance component including harmonic components and low resonant component including transient impact. The harmonic component is the narrow band signal and has high quality factor. On the contrary, the transient impact component is the broadband signal and has low quality factor. Thus, the harmonic component and transient impact component can be separated according to the difference of quality factor. By using resonance-based sparse signal decomposition method, the interference to fault diagnosis caused by the signal component which does not contain the fault characteristics can be avoided, and the reliability of fault diagnosis results can be assured.Supported by the project of Natural Science Foundation of China, which named as "Research on the resonance-based sparse signal decomposition method and its application in mechanical fault diagnosis (Project approval no:51275161)", this thesis researched the problem of gear fault diagnosis with variable rotating speed based on the resonance-based sparse signal decomposition. The main research works of the thesis are as follows:(1) According to the common types of gear faults and their corresponding characteristics, the vibration mechanism, the vibration signal characteristics and the commonly used gear fault diagnosis methods are introduced. (2) The resonance-based sparse signal decomposition method is analyzed. The signal resonance attribute, quality factor adjustable wavelet transform and the realization of separating high resonance component from low resonant component, which are the basis of the resonance-based sparse signal decomposition, are introduced. The resonance-based sparse signal decomposition method is applied to the fault vibration signal analysis of gears, such as the cracked gear and the broken gear. The results show that this method is an effective fault diagnosis method for gears.(3) Aimed at the difficulties to extract the fault characteristics from the vibration signal of a fault gear with variable rotating speed, an order domain analysis method based on resonance-based sparse signal decomposition is proposed. The resonance-based sparse signal decomposition method can separate the harmonic component and the transient impact component. The gear’s rotation speed obtained by chirp let path pursuit algorithm is used for the order domain analysis of the transient impact component. The proposed method can avoid the influences of harmonic component and guarantee the correctness of the transient impact component’s angle resampling result obtained by the extracted speed information. Therefore, the final result obtained by this method can reflect the gear fault information correctly. Application examples verify the effectiveness of the proposed method.(4) Aimed at the problem of early gear fault diagnosis with variable rotating speed, an order cyclo stationary demodulating method based on resonance-based sparse signal decomposition is proposed. Firstly, in order to get rid of the influences of harmonics components, the resonance-based sparse signal decomposition method is used to eliminate the harmonics components. Then, the order cyclostationary demodulating method is used for demodulation analysis of the transient impact component got from the vibration signal with variable rotating speed. The order cyclostationary demodulation spectrum of the transient impact component obtained by this method can reflect the fault information clearly. Application example of the early cracked fault diagnosis of a gear verifies its effectiveness.In this thesis, the resonance-based sparse signal decomposition method, which is suitable for extracting the fault information and separating interference component, is researched. The order domain analysis method and the order cyclostationary demodulating method based on the resonance-based sparse signal decomposition are proposed and applied to the fault diagnosis of gears with variable rotating speed. Simulation and application examples show that the resonance-based sparse signal decomposition method has a good application prospect in the fault diagnosis of gears with variable rotating speed.