Search

Rolling bearings are prone to failure due to the complexity and serious operational environment of rotating equipment. Intelligent fault diagnosis based on convolutional neural networks (CNNs) has become an effective tool to ensure the reliable operation of rolling bearings. However, interference caused by environmental noise and variable working conditions can affect the data. To solve this problem, we propose an improved fault diagnosis method called deep convolutional neural network based on multi-scale features and mutual information (MMDCNN). In our approach, a multi-scale convolutional layer is placed at the front end of a 1D_CNN to maximize the retention of the multi-scale initial features. Meanwhile, the key fault features are further enhanced adaptively by introducing a self-attention mechanism. Then, the composite loss function is constructed by maximizing mutual information as an auxiliary loss based on cross-entropy loss; thus, the proposed method can extract robust fault features with high generalization performance. To demonstrate the superiority of MMDCNN, we compared the performance of our scheme with several existing deep learning models on two datasets. The results show that the proposed model successfully achieves bearing fault diagnosis with interference from noise and variable working conditions, possessing a powerful fault feature extraction capability. [ABSTRACT FROM AUTHOR]

Imbalance is a typical characteristic of data in the field of intelligent fault diagnosis. As a data augmentation method that both balances data and extends information, the generative adversarial network has aroused widespread concern in the data imbalance problem. However, it still suffers from high difficulty in model training and poor quality of generated samples. To address this issue, this paper proposes a progressive generative adversarial network, which is a strategy able to gradually generate complex signals from low-frequency signals to wide-frequency signals. In the developed network, on the one hand, decomposing signals into different frequency bands explicitly reduces the difficulty of directly generating high-dimensional and wide-frequency signals. On the other hand, different generative models can focus on different frequency bands, avoiding the interference of high-energy signals on the fault feature frequencies. In order to achieve good generative performance, the generators and training method for the progressive generative adversarial network are elaborately designed. The correlation shortcut avoids information suppression due to the poor selection of noise amplitudes, and the focusing shortcut forces the body of the generator to focus on key feature bands. Ultimately, the superiority of the progressive generative network is verified by evaluating the quality of the generated samples on three datasets and applying the generated samples to solve the data imbalance problem. • A progressive generation framework is proposed to improve the quality of the generated signal. • The proposed generator retains the correlation between low and high frequencies. • The proposed generator can focus on specific frequency bands. • Model at each level is trained independently to avoid the accumulation of errors. [ABSTRACT FROM AUTHOR]