Bearing fault diagnosis based on multi-scale spectral images and convolutional neural network

• Published in: Journal of Vibroengineering
• Main Authors: Luo, Tongchao, Qiu, Mingquan, Wu, Zhenyu, Zhao, Zebo, Zhang, Dingyou
• Format: Journal Article
• Language: English
• Published: 24.08.2025
• ISSN: 1392-8716; 2538-8460
• DOI: 10.21595/jve.2025.24934

To address the challenges of poor performance in traditional diagnosis methods and two-dimensional (2-D) feature based approaches, this paper proposes a novel fault diagnosis approach based on multi-scale spectrum feature images and deep learning. Firstly, the vibration signal is preprocessed through mean removal processing and then converted to multi-length spectrum with fast Fourier transform (FFT). Secondly, a novel 2-D feature called multi-scale spectral image (MSSI) is constructed by multi-length spectrum paving scheme. Finally, a deep learning framework, convolutional neural network (CNN), is formulated to diagnose the bearing faults. Two experimental cases are utilized to verify the effectiveness of the proposed method. Experimental results demonstrate that the proposed method significantly improves the accuracy of fault diagnosis.