Fault diagnosis of a planetary gearbox based on a local bi-spectrum and a convolutional neural network

Abstract

The vibration signals of a planetary gearbox have the characteristics of strong background noise and instability and are non-Gaussian. Bi-spectrums can suppress Gaussian colored noise and are suitable for vibration signal processing of planetary gearboxes. In the traditional fault diagnosis methods based on bi-spectrums, the fault characteristic frequency amplitudes of bi-spectrum or bi-spectrum slices, or the further quantitative calculations of fault characteristic values, are generally used as the basis of fault diagnosis processes. It has been found that bi-spectrum images can directly characterize the faults of the planetary gearboxes. Convolutional neural networks (CNNs) have been used in mechanical fault diagnoses in recent years. One-dimensional original signals are converted into two-dimensional images as CNN input, which is an effective method for mechanical fault diagnoses. At the present time, there has not been any relevant research conducted using bi-spectral images as CNN input. In this study, a fault diagnosis method based on local bi-spectrum and CNN was proposed. A bi-spectral analysis of the vibration signals of the planetary gearbox was first carried out in order to reveal the fault information while retaining the non-Gaussian information. Then, according to the bi-spectrum symmetry, local images containing the entire domain information were taken as the input of the CNN, which reduced the redundancy of the fault information. Then, in order to improve the diagnostic accuracy of the CNN, the key parameters of CNN architecture were optimized. Finally, a CNN diagnosis model was built to realize the classification diagnoses of different fault positions and different fault degrees of planetary gearboxes. This study’s comparison of the diagnosis results of the full bi-spectrum+CNN, local bi-spectrum+SVM, original vibration signal+CNN, and local bi-spectrum+BP neural networks showed that the method proposed in this study had achieved both accuracy and rapidity in the fault diagnoses of planetary gearboxes.

A review on deep learning in planetary gearbox health state recognition: methods, applications, and dataset publication

“…Yang et al [98] adopted cyclic wavelet spectrum analysis to produce 2D images of planet bearing signals and fed them into an improved CNN model in which an extreme learning machine was adopted instead of a softmax classifier. Jiang et al [99] fed 2D bispectrum images into a CNN model to realize the fault diagnosis of a planetary gearbox. The 2D images generated by TFA methods or signal processing methods are obtained before they are fed into CNN models.…”

Support vector machine fault diagnosis based on sparse scaling convex hull

“…The data set selected for the experiment is shown in table 2, which is equivalent to a ten-class classification problem, which not only needs to classify the location of the fault, but also needs to determine its severity [35]. Figure 13 shows the data set distribution of its ten types of sample sets under three-dimensional (3D) features, and the labels in the figure correspond to the data set division 2.…”

A Bayesian CNN-based fusion framework of sensor fault diagnosis

“…Although there are gaps in research on utilizing deep learning methods for sensor fault diagnosis, current methods can still identify sensor faults thanks to advancements in deep learning techniques. As one of the popular deep learning algorithms, convolutional neural networks (CNN) has become a research topic in the fault diagnosis of electric motors [16,17], rolling bearings [18,19], and gearboxes [20,21]. Xu et al [22] proposed a global contextual residual convolutional neural network, which established a new hierarchical structure with a global context module, to solve the fault diagnosis of motors in variable-speed scenarios.…”