Abstract:Deep learning has been increasingly used in health management and maintenance decision-making for rotating machinery. However, some challenges must be addressed to make this technology more effective. For example, the collected data is assumed to follow the same feature distribution, and sufficient labeled training data are available. Unfortunately, domain shifts occur inevitably in real-world scenarios due to different working conditions, and acquiring sufficient labeled samples is time-consuming and expensive in complex environments. This study proposes a novel domain adaptive framework called deep Multiscale Conditional Adversarial Networks (MCAN) for machinery fault diagnosis to address these shortcomings. The MCAN model comprises two key components. Constructed by a novel multiscale module with an attention mechanism, the first component is a shared feature generator that captures rich features at different internal perceptual scales, and the attention mechanism determines the weights assigned to each scale, enhancing the model's dynamic adjustment and self-adaptation capabilities. The second component consists of two domain classifiers based on Bidirectional Long Short-Term Memory (BiLSTM) leveraging spatiotemporal features at various levels to achieve domain adaptation in the output space. The deep domain classifier also captures the cross-covariance dependencies between feature representations and classifier predictions, thereby improving the predictions' discriminability. The proposed method has been evaluated using two publicly available fault diagnosis datasets and one condition monitoring experiment. The results of cross-domain transfer tasks demonstrated that the proposed method outperformed several state-of-the-art methods in terms of transferability and stability. This result is a significant step forward in deep learning for health management and maintenance decision-making for rotating machinery, and it has the potential to revolutionize its future application.

Multiscale ECA network: a rotation mechanical domain adaptation method with minimal class confusion

Multiscale Dilated Convolutional Subdomain Adaptation Network with Attention for Unsupervised Fault Diagnosis of Rotating Machinery Cross Operating Conditions

Multi-Scale Attention Convolution Subdomain Adaption Network for Cross-Domain Fault Diagnosis of Machine

Multiscale Conditional Adversarial Networks based domain-adaptive method for rotating machinery fault diagnosis under variable working conditions

A Deep Convolution Multi-Adversarial adaptation network with Correlation Alignment for fault diagnosis of rotating machinery under different working conditions

A Rotating Machinery Fault Diagnosis Method Based on Multi-Sensor Fusion and ECA-CNN

Rotary Machinery Fault Diagnosis Based on Split Attention Mechanism and Graph Convolutional Domain Adaptive Adversarial Network

An Adaptive Weighted Multiscale Convolutional Neural Network for Rotating Machinery Fault Diagnosis under Variable Operating Conditions

Multiscale convolutional conditional domain adversarial network with channel attention for unsupervised bearing fault diagnosis

Multiscale Residual Attention Convolutional Neural Network for Bearing Fault Diagnosis

Multi-source domain adaptive network based on local kernelized higher-order moment matching for rotating machinery fault diagnosis

A convolutional multisensor fusion fault diagnosis framework based on multidimensional distance matrix for rotating machinery

Multiattention-Based Feature Aggregation Convolutional Networks with Dual Focal Loss for Fault Diagnosis of Rotating Machinery under Data Imbalance Conditions

A Cooperative Convolutional Neural Network Framework for Multisensor Fault Diagnosis of Rotating Machinery

Intelligent Cross-domain Fault Diagnosis For Rotating Machinery Using Multiscale Adversarial Convolutional Neural Network

A Multi-Scale Kernel-Based Network with Improved Attention Mechanism for Rotating Machinery Fault Diagnosis under Noisy Environments

Multiscale Representations Fusion with Joint Multiple Reconstructions Autoencoder for Intelligent Fault Diagnosis

An Unsupervised Rotating Machinery Fault Diagnosis Method Based on Multi-Scale Feature Residual Neural Network

TDMSAE: A Transferable Decoupling Multi-Scale Autoencoder for Mechanical Fault Diagnosis

Multibranch and Multiscale Dynamic Convolutional Network for Small Sample Fault Diagnosis of Rotating Machinery

Multi-representation Symbolic Convolutional Neural Network: a Novel Multisource Cross-Domain Fault Diagnosis Method for Rotating System