A fault diagnosis method for variable speed gearbox bearing based on SET improved multi-source ridge line. (15th June 2023)

• Record Type: Journal Article
• Authors: Wan, Shuting; Zhao, Xiaoyan; Wang, Yanjie; Zhang, Bolin; Zhang, Xiong; Gu, Xiaohui
• Is Part Of: Measurement. Volume 214(2023)
• Publication Date: 2023-06-15
• Subjects: Gearbox -- Multi-source time–frequency ridge line -- Tacho-less -- Order tracking -- Time–frequency analysis; Weights and measures -- Periodicals; Measurement -- Periodicals; Measurement; Weights and measures; Periodicals; 530.8
• DOI: 10.1016/j.measurement.2023.112758
• Languages: English
• ISSNs: 0263-2241

Abstract

Highlights: A SET-based improved multi-source ridge line fusion variable speed gearbox bearing fault diagnosis method is proposed. The method optimizes the accuracy of time–frequency ridge line based on time–frequency analysis. The method can improve the time–frequency ridge line frequency mutation phenomenon and optimize the bearing fault feature extraction. Both experimental and simulation results verify the effectiveness and superiority of the proposed method. Abstract: Due to the noise interference and frequency ambiguity of gearbox vibration signal under variable speed, the traditional instantaneous frequency estimation method is prone to frequency mutation and low ridge line fusion accuracy. This paper proposes an improved method of extracting multi-source time–frequency ridge lines based on time–frequency analysis by exploring the relationship between the vibration signal of the gearbox without a tachometer and the fluctuation law of the bearing rotation frequency. In the proposed method, the Variable Mode Decomposition (VMD) filter and Synchroextracting Transform (SET) are first used to analyze the time frequency of the gearbox vibration signal. Secondly, the local peak search algorithm obtains multiple time–frequency ridge lines from low-frequency and high-frequency resonance frequency bands. Thirdly, the synchronization process is carried out based on the maximum energy spectrum in the time spectrum. The multi-source ridge line extraction algorithm is improved with reduced frequency mutation and optimized ridge line fusion accuracy. Finally, the optimized speed curve is used to track the vibration signal order of the gearbox. The bearing fault characteristic order is obtained from the angular order spectrum, and the effectiveness of the proposed method is verified by the simulation and the experiment.