Application Research Of Tunable Q-factor Wavelet Transform On High-speed Train Gearbox Fault Diagnosis

In recent years,with the rapid development of China's high-speed railways,the speed of trains has been continuously improved,which puts forward higher requirements for the safety of train operation.The gearbox is the core component of the high-speed train transmission system.Once a fault occurs,it will directly threaten the safety of the train.The gearbox vibration signal contains most fault information of components in the gearbox,but the vibration signal composition is very complicated due to the complexity and harshness of its operating environment in high-speed trains.It is the focus of this paper to study how to effectively separate and decompose the gearbox vibration signal to achieve comprehensive gearbox composite fault diagnosis and prevent misdiagnosis and missed diagnosis.The Tunable Q-Factor Wavelet Transform(TQWT)can flexibly select the wavelet base to match the signals of different oscillation characteristics,at the same time,it has translation invariance,Therefore,it is possible to better deal with the fault signal mainly caused by the fault of the bearing and the broken teeth and spalling of the gear.Aiming at the problem of TQWT decomposition parameter selection and decomposition component evaluation,this paper proposes an IK index which can measure signal impact strength and periodicity simultaneously.Based on this,an adaptive TQWT method based on IK is proposed.The method adaptively searches for optimal parameters according to signal characteristics to decompose the signal,and selects the components with rich impact characteristics according to the component selection rule,then reconstructs these components to extract fault features in the signal to realize fault diagnosis.Through the processing of the simulated signal and the measured signal,the effectiveness and and advantages of the proposed method for the processing of shock-like fault characteristic signal are verified.On the basis of TQWT,Selesnick proposed resonance sparse decomposition,which can effectively separate the impact components and harmonic components in the signal according to the resonance property of the signal,and reduce the mutual interference and influence of the signal components well.Aiming at the problem that the decomposition effect of the method depends on the parameter selection,the paper introduces a global search optimization method—differential evolution method,and improves it.Combining the improved differential evolution method with the resonance sparse decomposition,the adaptive resonance sparse decomposition method is proposed.The method takes the maximum IK index of low resonance component as the optimization target to obtain the optimal decomposition parameters and the best decomposition effect of the signal.The effectiveness of the signal nonlinear separation of the proposed method is verified by the analysis and processing of the simulated signal and the application examples.After analyzing the composite fault condition of the gearbox,the gearbox composite fault vibration signal is divided into two types.Based on the advantages of the adaptive resonance sparse decomposition method and the adaptive TQWT method,Combine them and use the joint method in the processing of the gearbox composite simulation signal and gearbox composite fault detection which collected from the high-speed train gearbox tracking experiment.The method can uncover the fault characteristics of the hidden bearing while diagnosing the gear fault,and realize the fault diagnosis of the bearing.Therefore,the method proposed in this paper can effectively diagnose the composite fault of high-speed train gearboxes and avoid the occurrence of misdiagnosis and missed diagnosis.