Deep learning-based rolling bearing fault diagnosis method has made significant achievements, but its diagnostic performance is still limited by few samples. Aiming at this problem, a novel Intelligent Fault Diagnosis (IFD) method for rolling bearings, named Multi-scale Residual Anti-noise Network (MRANet) via interpretable Dynamic Recalibration Mechanism (DRM), is proposed. Firstly, the raw vibration signal is generated into a time-frequency diagram with more characteristic domains by Short-Time Fourier Transform (STFT). Then, the shallow mechanism and deep discriminable features are extracted using multi-branch dilated convolution and improved residual blocks. Simultaneously, the DRM assists the feature extractor to adaptively adjust the feature weights from the spatial position and the channel information ratio, respectively, to enhance the local impulse excitation. Furthermore, the corrective effect of DRM on the feature extractor is visualized, which improves the interpretability of the network. Comparative experiments are conducted with other popular IFD methods on public and Lanzhou University of Technology (LUT) bearing dataset, and the results show that MRANet can exhibit superior diagnostic performance with few samples under variable load and multi-speed conditions.
12 July 2022 in IEEE Transactions on Instrumentation and Measurement
Bearing is a key component in rotary machines, and the performance of the rotary machines mostly depends on the bearing health condition. In order to improve the safety and maintenance plan of the product based on the bearing condition, a monitoring indicator is constructed to identify the health condition of bearings in real time. First, the vibration signal is processed by the proposed maximal difference of the detail components in multiresolution singular value decomposition (MDDCs-MRSVD) algorithm. Second, the features matrix is constructed by selected features to reflect the health condition of bearings. Then, the gradient standard deviation of each sampling time is obtained by the gradient in the amplitude direction of the features matrix. Finally, a monitoring indicator can be constructed to identify healthy stages of bearing. The proposed methods are verified via the tested datasets provided by Intelligent Maintenance Systems, and Xi’an Jiaotong University and the Changxing Sumyoung Technology Company Ltd (XJTU-SY). The results indicate that the proposed method is efficient and accurate to monitor and identify the health stages of bearing in real time.