关键词: 滚动轴承, 故障诊断, 多尺度卷积神经网络, 混合注意力机制, 双向门控循环单元

Abstract: To address the problems of weak adaptive feature extraction capacity and poor diagnostic accuracy of conventional rolling bearing fault diagnosis methods, this paper proposes a deep learning-based fault diagnosis method combining a hybrid attention mechanism with a multi-scale convolutional neural network and bidirectional gated recurrent unit. The method uses convolution kernels of different sizes to capture multi-scale features of vibration signals. A hybrid attention mechanism is employed to assign weights to different parts of the feature sequence, enhancing feature representation. The bidirectional gated recurrent unit is used to extract the relationship between the features and enable layer-by-layer information propagation. The method is verified by experiments with different bearing datasets, and the results show that the accuracy of this method is 99. 86%, demonstrating the significant advantages of the proposed bearing fault diagnosis method in terms of feasibility and performance.

Key words: rolling bearings, fault diagnosis, multi-scale convolutional neural network, hybrid attention mechanism, bidirectional gated recurrent unit