A novel short-frequency slip fault energy distribution-based demodulation technique for gear diagnosis and prognosis. (3rd April 2020)

• Record Type: Journal Article
• Authors: Tong, Shuiguang; Huang, Yuanyuan; Tong, Zheming; Cong, Feiyun
• Is Part Of: International journal of advanced robotic systems. Volume 17:Number 2(2020:Mar./Apr.)
• Publication Date: 2020-04-03
• Subjects: Gearbox -- fault diagnosis -- prognosis -- RSSD -- SF-SFED graph; Robotics -- Periodicals; Robotics; Periodicals; 629.892
• DOI: 10.1177/1729881420915032
• Languages: English
• ISSNs: 1729-8806

Abstract

To conduct diagnosis and prognosis of gears, this paper introduces a novel short-frequency slip fault energy distribution-based demodulation method. As an essential step of the method, the resonance-based sparse signal decomposition algorithm is firstly employed to obtain the high-resonance part from the raw gear fault signal. To deal with the difficulty in determining the resonance frequency band, we establish a multi-input signal-output model to describe the signal components acquired from a faulty gear. Based on it, the short-frequency slip fault energy distribution graph is defined to locate the center frequency. Besides, the maximum amplitude in the short-frequency slip fault energy distribution graph can be used as a health indicator for prognosis, which is named as fault-induced resonance energy ratio. The effectiveness of the proposed method is validated with both simulated signal and test data. The positive results achieved in both experiments show the perfect property of the methodology for gear fault detection with high noise, especially when the fault is incipient. In addition, by comparing the fault-induced resonance energy ratio values of faulty gears with different severity, it is proved to be a reliable health indicator for gear prognostic.