Title :
A Novel Time-Frequency Analysis Method for Mechanical Failure Detection
Author_Institution :
Harbin Engineering University
Abstract :
This research will attempt to analyze the effectiveness of the Empirical Mode Decomposition (EMD) in diagnostics of gearboxes operating under non-stationary rotational speed. For this, a simple software program for signal processing and extraction of mechanical failure detection was developed and tested. The primary aim is to reveal the nonlinear behavior of the system by using time frequency methods as an alternative to Fourier analysis methodology. Both simulated and experimental response data are analyzed by applying EMD and time-frequency analysis. The following examples demonstrate the improvements in the detection of gear faults.
Keywords :
Acoustic signal detection; Acoustical engineering; Data analysis; Data mining; Failure analysis; Fault detection; Gears; Interference; Signal analysis; Time frequency analysis;
Conference_Titel :
Hybrid Information Technology, 2006. ICHIT '06. International Conference on
Conference_Location :
Cheju Island
Print_ISBN :
0-7695-2674-8
DOI :
10.1109/ICHIT.2006.253586