Using empirical mode decomposition scheme for helicopter main gearbox bearing defect identification

Abstract

Vibration sensors for helicopter health and condition monitoring have been widely employed to ensure the safe operation. Through the years, vibration sensors are now commonly placed on helicopters and have claimed a number of successes in preventing accidents. However, vibration based bearing defect identification remains a challenge since bearing defects signatures are usually contaminated by background noise resulting from variable transmission paths from the bearing to the receiving externally mounted vibration sensors. In this paper, the empirical mode decomposition (EMD) scheme was utilized to analyze vibration signal captured from a CS29 Category 'A' helicopter main gearbox, where bearing faults were seeded on one of the planetary gears bearing of the second epicyclic stage. The EMD scheme decomposed vibration signal into a number of intrinsic mode functions (IMFs) for subsequent envelope analysis. The selection of appropriate IMFs to characterize bearing fault signatures was discussed. The analysis result showed that the bearing fault signatures were successfully characterized and revealed the efficacy of the EMD scheme.

Radio Frequency Fingerprint Collaborative Intelligent Blind Identification for Green Radios

“…Duan et al proposed a method of using short-time Fourier transform to extract the time-frequency distribution of the signal as a fingerprint feature recognition method. This method was more suitable for non-stationary signals, such as chirp signals, but it was less effective for complex and irregular signals [6]. Reference [7] proposed an identification method based on the Wegener-Well distribution, but this method had the disadvantage of the existence of cross-interference terms.…”