A small sample size and unbalanced sample distribution are two main problems when data-driven methods are applied for fault diagnosis in practical engineering. Technically, sample generation and data augmentation have proven to be effective methods to solve this problem. The generative adversarial network (GAN) has been widely used in recent years as a representative generative model. Besides the general GAN, many variants have recently been reported to address its inherent problems such as mode collapse and slow convergence. In addition, many new techniques are being proposed to increase the sample generation quality. Therefore, a systematic review of GAN, especially its application in fault diagnosis, is necessary. In this paper, the theory and structure of GAN and variants such as ACGAN, VAEGAN, DCGAN, WGAN, et al. are presented first. Then, the literature on GANs is mainly categorized and analyzed from two aspects: improvements in GAN’s structure and loss function. Specifically, the improvements in the structure are classified into three types: information-based, input-based, and layer-based. Regarding the modification of the loss function, it is sorted into two aspects: metric-based and regularization-based. Afterwards, the evaluation metrics of the generated samples are summarized and compared. Finally, the typical applications of GAN in the bearing fault diagnosis field are listed, and the challenges for further research are also discussed.

Convolutional Neural Network (CNN) has been widely used in bearing fault diagnosis in recent years, and many satisfying results have been reported. However, when the training dataset provided is unbalanced, such as the samples in some fault labels are very limited, the CNN’s performance reduces inevitably. To solve the dataset imbalance problem, a Generative Adversarial Network (GAN) has been preferably adopted for the data generation. In published research studies, GAN only focuses on the overall similarity of generated data to the original measurement. The similarity in the fault characteristics is ignored, which carries more information for the fault diagnosis. To bridge this gap, this paper proposes two modifications for the general GAN. Firstly, a CNN, together with a GAN, and two networks are optimized collaboratively. The GAN provides a more balanced dataset for the CNN, and the CNN outputs the fault diagnosis result as a correction term in the GAN generator’s loss function to improve the GAN’s performance. Secondly, the similarity of the envelope spectrum between the generated data and the original measurement is considered. The envelope spectrum error from the 1st to 5th order of the Fault Characteristic Frequencies (FCF) is taken as another correction in the GAN generator’s loss function. Experimental results show that the bearing fault samples generated by the optimized GAN contain more fault information than the samples produced by the general GAN. Furthermore, after the data augmentation for the unbalanced training sets, the CNN’s accuracy in the fault classification has been significantly improved.

Convolutional Neural Network (CNN) has been extensively used in bearing fault diagnosis and Remaining Useful Life (RUL) prediction. However, accompanied by CNN’s increasing performance is a deeper network structure and growing parameter size. This prevents it from being deployed in industrial applications with limited computation resources. To this end, this paper proposed a two-step method to build a cell-based light CNN by Neural Architecture Search (NAS) and weights-ranking-based model pruning. In the first step, a cell-based CNN was constructed with searched optimal cells and the number of stacking cells was limited to reduce the network size after influence analysis. To search for the optimal cells, a base CNN model with stacking cells was initially built, and Differentiable Architecture Search was adopted after continuous relaxation. In the second step, the connections in the built cell-based CNN were further reduced by weights-ranking-based pruning. Experiment data from the Case Western Reserve University was used for validation under the task of fault classification. Results showed that the CNN with only two cells achieved a test accuracy of 99.969% and kept at 99.968% even if 50% connections were removed. Furthermore, compared with base CNN, the parameter size of the 2-cells CNN was reduced from 9.677MB to 0.197MB. Finally, after minor revision, the network structure was adapted to achieve bearing RUL prediction and validated with the PRONOSTIA test data. Both tasks confirmed the feasibility and superiority of constructing a light cell-based CNN with NAS and pruning, which laid the potential to realize a light CNN in embedded systems.

In data-driven bearing fault diagnosis, sufficient fault data are fundamental for algorithm training and validation. However, only very few fault measurements can be provided in most industrial applications, bringing the dynamics model to produce bearing response under defects. This paper built a Modelica model for the whole bearing test rig, including the test bearing, driving motor and hydraulic loading system. First, a five degree-of-freedom (5-DoF) model was proposed for the test bearing to identify the normal bearing dynamics. Next, a fault model was applied to characterize the defect position, defect size, defect shape and multiple defects. The virtual bearing test bench was first developed with OpenModelica and then called in Python with OMPython. For validation of the positive effect of the dynamics model in the direct transfer learning for bearing fault diagnosis, the simulation data from the Modelica model and experimental data from the Case Western Reserve University were fed separately or jointly to train a Convolution Neural Network (CNN). Then the well-trained CNN was transferred directly to achieve the fault diagnosis under the test set consisting of experiment data. Additionally, 157 features were extracted from both time-domain and frequency-domain and fed into CNN as input, and then four different validation cases were designed. The results confirmed the positive effect of simulation data in the CNN transfer learning, especially when the simulation data were added as auxiliary to experimental data, and improved CNN classification accuracy. Furthermore, it indicated that the simulation data from the bearing dynamics model could play a part in the actual experimental measurement when the collected data were insufficient.

Trajectory similarity-based evaluation is the most intuitive method for remaining useful life (RUL) prediction when abundant run-to-failure data are available. However, practical scenarios often present a challenge with limited access to such trajectories. This article introduces an improved similarity-based method to bridge the gap by employing Monte Carlo to generate more trajectories. To begin, 13 features are extracted from both time and frequency domains, subsequently conducting dimension reduction to build the health index (HI) corresponding to the original acceleration measurements. Afterward, HI trajectories are fit using exponential functions, and three different distribution functions (Gaussian, Gamma, Weibull) are adopted to identify the probability density of the two coefficients ( $\theta $ , $\beta $ ) in fit exponential models. Monte Carlo is applied to resample from the coefficient distribution, thereby generating more HI trajectories. Finally, the expanded HI library is used for RUL prediction and validated with two different bearing datasets. Experimental findings reveal that the trajectory expansion achieved through Monte Carlo sampling yields more accurate RUL estimation and reduced uncertainty.