Rolling bearings, as essential parts of rotating machinery, face significant challenges in fault diagnosis due to limited fault samples and high noise interference, both of which reduce the effectiveness of traditional methods. To tackle this, this study proposes a fault diagnosis approach that combines Digital Twin (DT) and deep learning. First, actual bearing vibration data were collected using an experimental platform. After denoising the data, a high-fidelity digital twin system was built by integrating the bearing dynamics model with a Generative Adversarial Network (GAN), thereby effectively increasing the fault data. Next, the Wavelet Synchro-Extracting Transform (WSET) is used for high-resolution time-frequency analysis, and convolutional neural networks (CNNs) are employed to extract deep fault features adaptively. The fully connected layer of the CNN is then combined with a Least Squares Support Vector Machine (LSSVM), with key parameters optimized through an Improved Pelican Optimization Algorithm (IPOA) to improve classification accuracy significantly. Experimental results based on both simulated and publicly available datasets show that the proposed model has excellent generalizability and operational flexibility, surpassing existing deep learning-based diagnostic methods in complex industrial settings.

Deep groove ball bearings are relatively weak in design for withstanding axial forces, but in practical applications, they may be subject to slight axial impact. Spalling failure may occur at a non-central position on the raceway. In response to this issue, this paper studies the axial impact characteristics generated in the bearing system due to uneven contact on both axial sides of the raceway when the rolling element passes through the defect area. A three-degree-of-freedom kinetic model considering the axial impact is proposed in this paper, simulating the effect of the impact on the rolling bearing when the flaking failure occurs at the non-central position of the roller path. The contact deformation area between the rolling element and raceway under axial forces and the position condition of the spalling fault in the raceway axial direction are introduced, and the impact characteristics of the spalling fault at different axial positions of the inner raceway are simulated by the time-varying energy method. Through comparative analysis of simulation and experiment, the variation characteristics of the vibration intensity in three directions of the bearing system when the spalling fault is at different axial positions of the raceway are obtained, and the correctness of the model is verified. This also provides theoretical support for improving the design and selection of bearings.