Fault Feature Extraction Of Flexible Thin-Walled Bearings Based On Spectral Kurtosis And MCKD

Harmonic reducer is an important precision reducer in industrial robot mechanical transmission system.As one of the core components of harmonic reducer,the stability of flexible thin-wall bearing determines the performance of harmonic reducer.Therefore,the research of vibration signal fault feature extraction for flexible thin-walled bearings is of great significance for the normal and stable operation of harmonic reducers.The vibration signal of the flexible thin-walled bearing is significantly different from that of the ordinary rolling bearing due to different structure and load characteristic.Aiming at the above problems,this paper takes the flexible thin-walled bearing as the research object,research the fault feature extraction method of the flexible thin-wall bearing vibration signal.The main research contents of the thesis are as follows:Firstly,by analyzing the working principle and kinematics of flexible thin-walled bearings,the theoretical calculation formula of the fault characteristic frequency of flexible thin-walled bearings is derived.The characteristics of vibration signals of flexible thin-walled bearings with no fault,inner ring fault and outer ring fault are analyzed.Then Fast Kurtogram algorithm and ASK(Adaptive Spectral Kurtosis)are introduced and applied to extract fault feature of flexible thin-wall bearing vibration signals.It's show that ASK method has better extraction effect.Since MCKD(Maximum Correlation Kurtosis Deconvolution)algorithm has the characteristics of extracting periodic shocks in strong noise background,The process of determining parameters of the MCKD algorithm for vibration signal of flexible thin-walled bearings is proposed.MCKD algorithm outperforms MED(Minimum Entropy Deconvolution)in extracting fault feature of flexible thin-walled bearings.Finally,analyzed the advantages and disadvantages of ASK method and MCKD algorithm,and proposed a method based on ASK-MCKD.By using the fault feature energy ratio in the envelope spectrum as the fault feature extraction evaluation standard,the advantage of ASK-MCKD method in fault frequency extraction of vibration signal of flexible thin-walled bearings is demonstrated.