K-SVD-based WVD enhancement algorithm for planetary gearbox fault diagnosis under a CNN framework

Abstract

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.

Fault diagnosis of rotating machines based on EEMD-MPE and GA-BP

“…The data-driven methods based on time-frequency analysis are also an effective tool for extracting mechanical health information from nonstationary signals [6]. Traditional time-frequency analysis methods, such as wavelet transform [7], Wigner-Ville distribution [8], and short-time Fourier transform [9], have made certain achievements in bearing fault diagnosis. However, these methods lack self-adaptive capability when processing bearing vibration signals for fault diagnosis.…”

Research on Fault Feature Extraction of Hydropower Units Based on Adaptive Stochastic Resonance and Fourier Decomposition Method

“…Fan and Zhu [28] used EMD and SVD to process vibration signals and achieved a satisfied effect in removing noise interference. Considering the unstable operating condition in industrial applications and the noise, Li et al [29] put forward a new method for planetary gear box fault diagnosis, using the singular value decomposition (SVD) to enhance the resolution of the TFD obtained by the Wigner-Ville distribution (WVD). e results showed that the proposed method can not only reduce the influence of the cross terms of WVD but also eliminate noise and improve the accuracy of fault diagnosis.…”

A method for degradation features extraction of diesel engine valve clearance based on modified complete ensemble empirical mode decomposition with adaptive noise and discriminant correlation analysis feature fusion

“…Because the vibration signal of the cylinder head is affected by noise and other disturbances, before extracting the degraded features, we need to preprocess the source signal to restore the actual signal. Nowadays, Wigner–Ville distribution (Li et al, 2020), wavelet transform (El-Hendawi and Wang, 2020), and empirical mode decomposition (EMD) (Cai et al, 2020) are commonly used filtering methods. Different from other methods because its basic function is decomposed by the data itself, empirical mode decomposition (EMD) has the advantages of intuitive, direct, posterior, and adaptive.…”