12 January 2026 in Mechanical Systems and Signal Processing
9 June 2025 in IEEE Transactions on Instrumentation and Measurement
Due to the coupling and diversity of compound fault features, it is difficult to decoupling diagnose the compound fault under small datasets and variable working conditions. To enhance the compound fault diagnosis performance, this work proposed a compound fault transfer diagnosis method that uses the multi-label transformer-based transfer learning network. Initially, the 1-D signal is converted into 2-D image using CWT as the input of the constructed deep learning network; Subsequently, a CNN-Transformer network structure is constructed, which simultaneously utilizes convolutional operation and self-attention mechanism to comprehensively capture the multi-scale fault features. In addition, multi-label classification is adopted to identify specific single fault types in the compound fault and output corresponding fault labels. Finally, a transfer learning strategy with fine-tuning mechanism is creatively integrated to achieve the compound fault transfer diagnosis, which can learn fault knowledge from single fault samples in source domain for compound fault diagnosis in required target tasks. The experimental datasets of planetary gearbox and harmonic reducer were collected in our own laboratory to validate the diagnostic performance. The results show that our proposed method achieved 98% average decoupling identification accuracy in insufficient compound fault samples on both datasets, and the accuracy fluctuation was less than 1.5% in the six transfer tasks. It is verified that the proposed method achieved excellent decoupling identification accuracy and generalization ability in small datasets and variable working conditions, which has significant advantage over the traditional local convolution-based multi-label classification methods and domain adaptation -based transfer learning methods.
31 May 2025 in Measurement
6 July 2026 in SSRN Electronic Journal
Harmonic reducers are key components in precision industrial robots, and their faults directly affect motion accuracy and reliability. Transmission error (TE) signals provide a direct and sensitive basis for early fault identification; however, multi-source coupling and weak fault features make complex fault representation challenging, thereby limiting the classification performance of diagnostic models. In this study, a TE-oriented diagnostic method is developed by integrating adaptive MPIM-FMD decomposition and Fusion-SwinT multi-mode feature fusion. First, a multi-dimensional periodic impulse metric (MPIM) is constructed by considering impulsiveness, periodicity, and modal redundancy, and MOPSO is used to optimize the key FMD parameters for enhanced weak and coupled fault feature extraction. Then, the selected fault-related modes are encoded into MGAF and MSTFT image representations to capture global temporal correlations and local time–frequency evolution. Furthermore, DFEM and CMFM are introduced to enhance discriminative intra-modal features and promote cross-modal feature interaction. The fused representations are fed into the Swin Transformer for final fault classification. Experimental results on the self-developed harmonic reducer TE dataset confirm that the proposed method can accurately identify both single and compound faults under various operating conditions.
17 May 2026 in Measurement