HILBERT-HUANG TRANSFORM BASED GEARBOX FAULT DIAGNOSIS

摘要

Demodulation and non-stationary modulating signals are two important issues in gearbox fault detection. Hilbert-Huang transform provides an effective way to address these issues. This paper reports an investigation on gearbox fault diagnosis using Hilbert-Huang transform. Hilbert-Huang transform decomposes signals using empirical mode decomposition into intrinsic mode functions and applies Hilbert transform to these functions to obtain the Hilbert spectrum. In this paper, real vibration signals collected from a gearbox dynamics simulator with one driving gear fault is used to evaluate the effectiveness of Hilbert-Huang transform. A methodology for extracting condition indicators using Hilbert-Huang transform is presented. Discussion on the influences of the loads on the effectiveness of the Hilbert-Huang transform based condition indicator for gearbox fault diagnosis is included in the paper.