Application of Transformer Based on Escape Optimization Algorithm to the Faulty Bearing Classification Problem

Zhangqi Song; Chenlu Zhao

doi:10.54097/01avec07

Authors

Zhangqi Song
Chenlu Zhao

DOI:

https://doi.org/10.54097/01avec07

Keywords:

Escape Optimization Algorithm, Bearing Fault Detection, Transformer, Deep Learning, Feature Extraction.

Abstract

Deep learning has shown strong feature extraction and classification abilities in fault diagnosis, especially for vibration signal time series. The Western Reserve bearing dataset is a common benchmark for evaluating diagnostic models. This paper proposes a Transformer model optimized with the Escape Algorithm (ESC) to enhance bearing fault feature extraction and classification accuracy. Innovations include a multi-scale convolution-attention fusion module for better local temporal feature extraction and the first use of ESC for Transformer hyperparameter optimization. Experiments on the Western Reserve dataset show the model achieves 96% accuracy on both training and test sets, with improved robustness in distinguishing ambiguous categories. The results demonstrate the model’s effectiveness for complex fault recognition, offering valuable insights for intelligent manufacturing and equipment health monitoring.

Downloads

Download data is not yet available.

References

[1] Chen S, Chen W H, Zhang W G. Research on bearing fault diagnosis method based on DRSN fusion Transformer encoder [J]. Automation and Instrumentation, 2024, 39 (5): 103–108.

[2] Fang K L, Xu H R, Liu W J. Fault diagnosis and measurement of rolling bearing cage displacement [J]. Sensor Technology and Application, 2022, 10 (3): 420–425.

[3] Tang D Q. Research on bearing remaining life prediction algorithm based on coupled diffusion [D]. Anhui: Anhui University, 2024.

[4] Liu X. Research on composite fault diagnosis of rolling bearing based on Transformer model [D]. Beijing: Beijing University of Chemical Technology, 2023.

[5] Qiu D W, Liu Z C, Zhou Y C, et al. Identification of rolling bearing fault types based on Transformer neural network [J]. High Technology Newsletter, 2021, 31 (1): 1–11.

[6] Chen W Q, An D, Zhang L, et al. A bearing fault diagnosis method based on linear transformer [J]. Coal Mine Machinery, 2024, 45 (7): 177–179.

[7] Zhao J, Li J, Li Z, et al. Transformer network enhanced by dual convolutional neural network and cross-attention for wheelset bearing fault diagnosis [J]. Frontiers in Physics, 2025, 13: 1546620.

[8] Zhao X, Zhao Z, Cui X, et al. Few-shot bearing fault diagnosis using adaptive detail convolution and global KAN-Transformer with Mahalanobis distance [J]. Journal of Vibration Engineering & Technologies, 2025, 13 (5): 296.

[9] Ouyang K, Fu S, Chen Y, et al. Escape: an optimization method based on crowd evacuation behaviors [J]. Artificial Intelligence Review, 2025, 58: 19.

[10] Yang J, Han H, Dong X, et al. Bearing fault diagnosis grounded in the multi-modal fusion and attention mechanism [J]. Applied Sciences, 2025, 15 (3): 1531.