An Adaptive Domain Adaptation Method for Rolling Bearings’ Fault Diagnosis Fusing Deep Convolution and Self-Attention Networks

Due to the frequent switch of the working condition, fault diagnosis model for rotating machinery established on the training set (the source domain) cannot be effectively applied to the test set (the target domain). Traditional domain adaptation methods address this issue by performing feature alignment between the source and target domain, which ignores the positive guidance provided by prior knowledge. In this paper, we propose a prior knowledge-driven domain adaptation (PKDA) method for varying working condition fault diagnosis of rotating machinery, where a self-supervised learning framework is designed to integrate expert prior knowledge and structural prior knowledge. First, an expert prior knowledge guidance module is designed to extract features with physical significance. Then we align the features between the source and target domain by the max mean discrepancy (MMD) metric. Besides, a progressive Shannon entropy minimization strategy is proposed to realize the feature distribution structure of intra-class compression and inter-class separation, which can effectively integrate the structural prior knowledge. In this way, PKDA can effectively utilize the prior knowledge to achieve better performance in varying working condition fault diagnosis tasks. The effectiveness of the proposed method is illustrated by an open-source rolling bearing fault dataset from Case Western Reserve University and an open-source gearbox fault dataset from Southeast University.

Section: Methodsmentioning

confidence: 99%

Section: Varying Working Condition Fault Diagnosis Taskmentioning

confidence: 99%

Section: Tests and Analyzationsmentioning

confidence: 99%

See 1 more Smart Citation

Knowledge-driven domain adaptation strategy for rotating machinery fault diagnosis under varying working condition

Chang,

Yao,

Chen

et al. 2024

“…Fault diagnosis methods mainly include vibration signalbased methods [6][7][8][9] and thermal signal-based methods [10,11]. The vibration signal has high accuracy, it is the most accessible signal in research and applications.…”

Section: Introductionmentioning

confidence: 99%

Fault diagnosis of bearing-rotor system based on infrared thermography: ReSPP with multi-scaled training method

An,

Liu,

Shao

et al. 2023

The fault diagnosis method of bearing-rotor system based on infrared thermography can reflect the global fault information of the equipment, which is an advanced non-contact monitoring measure. Current diagnosis methods focus on the analysis of single-scale input images, and the network only accepts fixed-size inputs. The model is not capable of fusing multi-scaled fault features while processing single-scaled images, and compression for the size constraint will cause geometric distortion. This paper proposed a novel fault diagnosis model ReSPP (Residual Network with Least Spatial Pyramid Pooling (LSPP)) based on the improved LSPP. LSPP solves the problem of fault feature distortion caused by fixed-size constraint of the network. By replacing the single-scaled training with the proposed multi-scaled training method applied to fault diagnosis, the weight parameters of ReSPP pool the deep fault features of the bearing-rotor system at multiple scales, retaining the critical fault semantic information. The proposed model solves the fault feature loss in the process of feature extraction by improving the subsampling residual block. Experimental results show that ReSPP with multi-scaled training method (ReSPP-MSTM) classifies the fault conditions of the bearing-rotor system with an average diagnostic accuracy of 99.18%.

“…Methods such as domain adversarial neural networks (DANNs) [19] and domain adaptive networks (DANs) [20] have achieved promising results. Yu et al [21] proposed a method that combines DANN with a multi-kernel maximum mean difference (MK-MMD) to construct a joint loss function for diagnostic tasks under variable working conditions. Xiang et al [22] transferred simulated data to actual working conditions for fault diagnosis and utilized MK-MMD to minimize the distributional difference between the source and target domains.…”

Section: Introductionmentioning

confidence: 99%

Unsupervised rolling bearing fault diagnosis method across working conditions based on multiscale convolutional neural network

Fu,

Yu,

Zhan

et al. 2023

In practical engineering, the features of rolling bearing vibration signals often vary in distribution under different working conditions, and obtaining sample labels for target working conditions can be challenging. Therefore, a multi-scale depth subdomain adaptive convolutional neural network (MS-DSACNN) fault diagnosis method is proposed. The MS-DSACNN method is based on a multi-scale feature extraction method combined with an improved 1D-ConvNeXt architecture, which fuses low- and high-level features into multi-scale fault features. A channel attention mechanism is also introduced to adaptively assign weights to the fused multi-scale features. The local maximum mean discrepancy is constructed by extracting features and their corresponding source and pseudo-label inputs for the source and target working conditions, respectively, to construct the loss function for backpropagation. The MS-DSACNN model was evaluated in experiments with two different loads and one variable speed dataset. The experimental results indicate that MS-DSACNN has higher feature extraction capability and diagnostic accuracy than other unsupervised cross-condition fault diagnosis techniques. Additionally, the scenario where the signal captured by the sensor contains a considerable amount of noise in actual working conditions is considered.