This paper presents a new method of planetary gearbox fault diagnosis by dealing with and analyzing vibration signals. This study contributes to the realization of automatic diagnosis using a convolution neural network (CNN) to process time-frequency distributions (TFDs) transformed from vibration time series. In order to solve the problem of non-stationary working states and strong noise interference in industrial applications, a K-singular value decomposition (K-SVD) is used to enhance the resolution of TFDs obtained by Wigner–Ville distribution (WVD), a typical time-frequency transform algorithm. The simulation results indicate that K-SVD can not only reduce the effects of cross-terms on WVDs but can also eliminate noise, which makes the fault characteristics outstanding in the time-frequency domain. The enhanced WVDs improve the accuracy of fault diagnosis in a classification framework based on the CNN that can extract features adaptively and obtain a high degree of discrimination between different fault conditions. Finally, the effectiveness of the proposed method is verified by a prototype experiment with roller bearings and a scale test rig of a planetary gearbox from a ship unloader. Moreover, a priority confusion matrix is proposed as a visualization tool with which to evaluate the performance of a fault diagnosis model. The results open the possibility of extrapolating the method to the fault diagnosis of other mechanical parts.
Weak fault detection is a complex and challenging task when two or more faults (compound fault) with discordant severity occur in different parts of a gearbox. The weak fault features are prone to be submerged by the severe fault features and strong background noise, which easily lead to a missed diagnosis. To solve this problem, a novel diagnosis method combining muti-symplectic geometry mode decomposition and multipoint optimal minimum entropy deconvolution adjusted (MSGMD-MOMEDA) is proposed for gearbox compound fault in this paper. Specifically, different fault components are separated by the improved symplectic geometry mode decomposition (SGMD), namely, multi-SGMD (MSGMD) method. The weak fault features are enhanced by the multipoint optimal minimum entropy deconvolution adjusted (MOMEDA). In the process of research, a new scheme of selecting key parameters of MOMEDA is proposed, which is a key step in applying MOMEDA. Compared with SGMD, the proposed MSGMD has two main improvements, including suppressing mode mixing and preventing the generation of the pseudo components. Compared with the original method of selecting parameters based on multipoint kurtosis, the proposed MOMEDA parameters selecting scheme has more merits of high accuracy and precision. The analysis results of two cases of simulation and experiment signal reveal that the MSGMD-MOMEDA method can accurately diagnose the gearbox compound fault even under strong background noise.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.