Planetary Gearbox Fault Diagnosis under Time-variant Conditions Based on Iterative Generalized Synchrosqueezing Transform

Abstract

The synchrosqueezing transform can effectively improve the readability of time-frequency representation of mono-component and constant frequency signals, but blurs still occur to the time-frequency representation when analyzing multi-component and time-variant frequency signals. In order to address the time-frequency blur issue, an improved synchrosqueezing transform based on iterative generalized demodulation is proposed. The signal is decomposed into mono-components of constant frequency, via iterative generalized demodulation and filtering. The instantaneous frequency and time-frequency representation of such mono-component are accurately estimated by synchrosqueezing transform. The time-frequency representation of original signal is obtained by superposing the reconstructed time-frequency representations of all the mono-components. This proposed method effectively improves the time-frequency readability of synchrosqueezing transform when analyzing time-variant frequency signals, and generalizes the synchrosqueezing transform to multi-component signals as well. Using this method, the vibration signal frequency contents and their time-varying characteristics of a wind turbine planetary gearbox under time-variant conditions are effectively identified, and the gear fault is diagnosed.