Compound Fault Diagnosis for Gearbox based on Variational ModeDecomposition and Correlated Kurtosis

Abstract

Aiming at the characteristics of non-linear, non-stationary, multi-noise and multi-fault signals of the compound fault signal in gearboxes, a compound fault diagnosis method for gearboxes based on variational mode decomposition and correlated kurtosis is proposed. According to the characteristic that the signal can be effectively decomposed into several compact intrinsic mode functions by variational mode decomposition, the compound fault signal are decomposed into several intrinsic mode functions which contains different faults. Then，according to the characteristics that correlated kurtosis can identify the impact signals with different periods, the intrinsic mode functions containing different faults are extracted by using the maximum correlated kurtosis principle. Finally, the envelop analysis of these intrinsic mode function is carried out to achieve the purpose of compound fault diagnosis. The validity of the proposed method is proved by the analysis of the experimental signal. In addition, the influence of the number of intrinsic mode functions on the result of the proposed method is analyzed, which provides an empirical conclusion for the follow-up study.

Keywords

• Gearbox