25 February 2025 in IEEE Transactions on Instrumentation and Measurement
Fault diagnosis of rotating machinery is widely recognized as a challenging problem. Recent advances in combining convolutional neural networks (CNNs) and Transformers have expanded the capabilities of intelligent fault diagnosis. However, in real-world industrial settings, three critical challenges substantially affect fault diagnosis performance: resource-constrained deployment environments, variable operating conditions, and cross-domain adaptability requirements. These challenges frequently undermine the efficacy of existing algorithms, particularly in sustaining robust performance while meeting lightweight implementation requirements. To address these challenges, a novel lightweight fault diagnosis framework, termed ARE-MGCAFormer, is introduced in this paper. Firstly, an Adaptive Residual Enhancement (ARE) block is specifically designed to extract multi-local receptive field features from vibration signals. The integrated gating mechanism utilizes learnable parameters to adaptively balance the contributions of deep separable convolutions and inverse residual blocks, dynamically adjusting to varying signal features and noise levels while maintaining a lightweight design. Secondly, a Multi-Group Coordinate Attention (MGCA) mechanism is incorporated to effectively extract critical detailed features across the entire signal range while reducing computational complexity by distributing attention across multiple feature groups. Experimental verification was conducted using the gearbox dataset from Xi’an Jiaotong University (XJTU) and the rolling bearing fault dataset from the University of Ottawa (OU). The results demonstrate that the proposed framework exhibits superior lightweight characteristics and robustness in fault diagnosis tasks compared to recent mainstream frameworks based on CNNs and Transformers.
11 December 2025 in IEEE Transactions on Instrumentation and Measurement
As critical components of rotating machinery, bearings play a vital role in mechanical systems, making their health management and maintenance highly significant. However, current data-driven fault diagnosis techniques for bearings face three major challenges: difficulty in extracting fault features under variable-speed conditions, challenges in isolating compound fault characteristics, and insufficient data protection mechanisms. To address these issues, this paper proposes a Secure Federated Learning framework integrating a Dual-channel Convolutional Neural Network with a Transformer (SFLDFormer), which enables effective extraction of compound fault features under variable speeds and achieves accurate fault diagnosis. In terms of data protection, a noise-injection-based secure aggregation mechanism is introduced to effectively defend against data reconstruction attacks during the early stages of federated learning, ensuring the privacy and security of rotating component fault data throughout the distributed modeling process. Experimental results demonstrate that the proposed method maintains high diagnostic accuracy for compound faults under both constant and variable speed conditions. This work paves a feasible path toward building a secure and trustworthy cross-enterprise collaborative diagnostic system.