Health indicator extraction based on sparse representation of vibration signal for planetary gearbox

Abstract

The general methods of health indicators extraction for planetary gearbox are based on the vibration signals which acquired by sensors equipped on the casing of the gearbox and sampled in the frame of Shannon sampling theory. Therefore, it is necessary to sample and save abundant original vibration data in the process of uninterrupted monitoring, and this will generate masses of original data which would burden the storage and transmission. For this issue, a health indicator extraction method based on sparse representation and reconstruction theory is proposed in this paper. It only needs to sample and save fewer compressive measurements of vibration signal directly compared to original signal. There is no need to recover the original signal accurately for extract health indicators, while it just requires some sparse representation and reconstruction results based on the redundant dictionary of the original signals. The effectiveness of the method proposed is validated with simulation data.