Accurate bearing fault diagnosis is of great significance of the safety and reliability of rotary mechanical system. In practice, the sample proportion between faulty data and healthy data in rotating mechanical system is imbalanced. Furthermore, there are commonalities between the bearing fault detection, classification, and identification tasks. Based on these observations, this article proposes a novel integrated multitasking intelligent bearing fault diagnosis scheme with the aid of representation learning under imbalanced sample condition, which realizes bearing fault detection, classification, and unknown fault identification. Specifically, in the unsupervised condition, a bearing fault detection approach based on modified denoising autoencoder (DAE) with self-attention mechanism for bottleneck layer (MDAE-SAMB) is proposed in the integrated scheme, which only uses the healthy data for training. The self-attention mechanism is introduced into the neurons in the bottleneck layer, which can assign different weights to the neurons in the bottleneck layer. Moreover, the transfer learning based on representation learning is proposed for few-shot fault classification. Only a few fault samples are used for offline training, and high-accuracy online bearing fault classification is achieved. Finally, according to the known fault data, the unknown bearing faults can be effectively identified. A bearing dataset generated by rotor dynamics experiment rig (RDER) and a public bearing dataset demonstrates the applicability of the proposed integrated fault diagnosis scheme.

The rolling bearing fault signals are complex time series with complex dynamic characteristics and non-uniform periodicity due to the influence of random interference, such as random impulse noise and equipment vibration. This will affect the accuracy of the diagnostis method. This paper analyses the characteristics of bearing fault signals and discussed the basic ideas of current diagnosis methods. In order to decrease the computational complexity in time series analysis process and reduce the influence of random interference in subsequent feature extraction, this paper proposes a rolling bearing fault diagnosis method base on periodic sparse attention and LSTM (PSAL) for non-uniform bearing vibration signals. According to the periodic characteristics of bearing fault, a periodic sparse attention network is proposed to decrease the time consumption in the process of reducing the influence of random interference and enhancing the feature. Then, LSTM is used to extract long-term dependence features in the fault signals. Finally, two sets of rolling bearing datasets are adopted to verify the validity and superiority of the proposed method by comparing with other methods.

Unsupervised domain adaptation is widely used for fault diagnosis under variable working conditions. However, loss oscillation and slow convergence, which are caused by the dynamically varying alignment of targets during domain adaptation, are ignored. Therefore, in this article, a Gaussian mixture variational-based transformer domain adaptation (GMVTDA) fault diagnosis method is proposed. A feature extractor based on transformer layers is designed to capture long-term dependence information and local features. Subsequently, a domain alignment term is proposed to project the features learned from both working conditions into the common assistance distribution and make them follow the same distribution after the alignment process. In addition, considering that fault diagnosis is a multiclassification process, a Gaussian mixture is utilized to build the common assistance distribution. Ultimately, the proposed GMVTDA is applied to bearing fault diagnosis under variable working conditions, and the experimental results prove its effectiveness.

The application of deep learning-based rolling bearing fault diagnosis methods in high reliability scenarios is limited due to low transparency. In addition, the scaling up of the deep learning models, in order to improve the performance of rolling bearing fault diagnosis (RBFD), has led to difficulties in its application in low-resource scenarios. Based on these facts, a new neural network, restricted sparse networks (RSNs), is proposed in this work. Firstly, a restricted sparse frequency domain space (RSFDS) is proposed for the interpretable representation of rolling bearing fault features (RBFFs) based on the quadratic complex domain equation. Secondly, an interpretable multi-channel fusion mechanism is designed to map RBFFs to RSFDS. Furthermore, a high power feature extraction module is developed to extract RBFFs in an efficient and easy-to-understand manner. Finally, an end-to-end RBFD network is provided for high reliability and resource-constrained scenarios. The experimental results show that RSNs have favorable fault diagnosis accuracy performance that is parallel to the state-of-the-art methods. More importantly, the model size of the proposed network only accounts for 20%-30% of the conventional methods.

Lack of massive-labeled samples may cause the performance degradation of intelligent bearing fault diagnosis methods under variable working conditions. Unsupervised domain adaptation (UDA)-based methods can effectively alleviate this problem by decreasing the distribution discrepancy between two dataset. However, most current UDA-based methods focus only on minimizing the marginal distribution. The influence of misclassification of samples near the classification boundaries is ignored. This may lead to the result that even though the marginal distributions are aligned well, the diagnosis accuracy is not ideal enough. On account of this, an unsupervised contrastive domain adaptation network (UCDAN) for bearing fault diagnosis under variable working conditions is proposed. A deep convolution feature extractor is constructed to extract high-level features. At the same time, a domain-adversarial total loss function based on the strategy of minimizing the domain distribution discrepancy is designed to obtain domain-shared discriminant features for aligning the marginal distribution. Moreover, a contrastive estimation term is designed to quantize the similarity of data distribution and maximize the consistency between the samples from the same health conditions. Thereby reducing the probability of samples being misclassified near or on the boundary and improving diagnosis performance. And information entropy is used to avoid most samples are assigned to the same cluster. The effectiveness of UCDAN is confirmed by examining various fault diagnosis situations with domain discrepancies across the source and target conditions, using experimental data from two bearing systems.