The increase in the number of channels for extracting bearing fault features can to some extent enhance diagnostic performance. Therefore, this article proposes a SENet (squeeze and excitation network)—TSCNN (two flow convolutional neural network) model with high accuracy and generalization characteristics for fault diagnosis of rolling bearings. Firstly, use convolutional pooling layers to construct a basic diagnostic model framework. Secondly, due to the unsatisfactory performance of feature extraction solely on one-dimensional frequency domain signals or two-dimensional time-frequency signals, there may be misjudgments. Therefore, a dual stream convolutional model is integrated to process both one-dimensional and two-dimensional data. Fast Fourier transform is used to process one-dimensional frequency domain data, and continuous wavelet transform is used to process two-dimensional time-frequency maps. Once again, integrating the SENet module into the dual stream diagnostic model, the addition of attention mechanism can enable the model to better understand key features of input data. Finally, the data obtained from the processing of two channels is fused and classified in the Softmax layer. This article uses the rolling bearing fault standard data from Case Western Reserve University and the American Society for Mechanical Fault Prevention Technology, and verifies through multiple controlled experiments that the model established in this article has high accuracy and good generalization characteristics.

According to the influence of the normal contact damping of joint surfaces on the dynamic characteristics of high-precision machinery (machine tool, robot, etc.), in this article, a three-dimensional fractal model of normal contact damping of dry-friction rough joint surfaces based on Hertz theory and fractal theory is established. The three-dimensional surface topography is constructed, according to the modified double variable Weierstrass–Mandelbrot function. The fractal model of strain energy $E_{\mathrm{e}}$ , dissipated energy $E_{\mathrm{p}}$ , damping loss factor $\mathrm{\Delta }$ , and normal contact damping $C_{\mathrm{n}}$ are deduced in detail, and the influence of fractal parameters and dynamic friction coefficient $\mu$ on them are simulated. The simulation curves show that strain energy $E_{\mathrm{e}}$ , dissipated energy $E_{\mathrm{p}}$ , damping loss factor $\mathrm{\Delta }$ , and normal damping $C_{\mathrm{n}}$ increase with the increase in the fractal roughness G; the influence of fractal dimension D on $E_{\mathrm{e}}$ is more changeable, first $E_{\mathrm{e}}$ decreases with the increase in D and then increases. $E_{\mathrm{p}}$ , $\mathrm{\Delta }$ , and $C_{\mathrm{n}}$ increase with the increase in D and $\mu$ , respectively; the effect of $\mu$ on $E_{\mathrm{e}}$ is not obvious, so simple change in $\mu$ has no significant change in $E_{\mathrm{e}}$ ; a comparative analysis of the theoretical calculation of normal damping and experimental results show that their general trend is consistent, and they increase with the increase in total normal contact load P, the relative error is 5%–25%. The theoretical model can provide reference for the design of normal contact damping of the joint surfaces.

The contact damping between rough surfaces has an important influence on the wear, vibration, contact fatigue, and energy dissipation between interfaces. In this paper, based on contact theory, a tangential damping mathematical model of rough surfaces is established from the point of view of viscous contact damping energy dissipation mechanism of asperities and considering the fractal characteristics of three-dimensional topography of rough surfaces. Through the combination of micro-contact modeling and macro dynamic testing of composite beams, the analysis results show that there are important evolution rules between tangential damping and surface fractal parameters and material parameters. The nonlinear relations between them are as follows: tangential contact damping is positively correlated with normal load, load ratio, and maximum contact area of asperity, and negatively correlated with fractal roughness; tangential contact damping increases first and then decreases with the increase of three-dimensional fractal dimension. The results of computational and experimental modal analysis show that the established mathematical model is feasible for predicting tangential damping. The study of tangential contact damping between surfaces can lay a foundation for improving the performance of assembly interfaces.

Research aimed at diagnosing rolling bearing faults is of great significance to the health management of equipment. In order to solve the problem that rolling bearings are faced with variable operating conditions and the fault features collected are single in actual operation, a new lightweight deep convolution neural network model called FC-CLDCNN, composed of a convolution pooling dropout group with two-stream feature fusion and cross-load adaptive characteristics, is proposed for rolling bearing fault diagnosis. First, the original vibration signal is transformed into a one-dimensional frequency domain signal and a two-dimensional time-frequency graph by fast Fourier transform and continuous wavelet transform. Then, the one-dimensional frequency domain signal and two-dimensional time-frequency diagram are input into the two channels of the model respectively to extract and recognize the one-dimensional and two-dimensional features. Finally, the one-dimensional and two-dimensional features are combined in the fusion layer, and the fault types are classified in the softmax layer. FC-CLDCNN has the characteristics of two-stream feature fusion, which can give full consideration to the characteristics of rolling bearing fault data, so as to achieve efficient and accurate identification. The Case Western Reserve University (CWRU) dataset is used for training and testing, and it is proved that the proposed model has high classification accuracy and excellent adaptability across loads. The Machinery Failure Prevention Technology (MFPT) dataset was used to validate the excellent diagnostic performance and generalization of the proposed model.