A time–frequency ridge extraction diagnostic method for composite faults of bearing gears in wind turbine gearboxes

Abstract

The bearing and gear faults in the gearbox interact with each other, and the weak fault characteristics are often masked by the strong fault characteristics, making it difficult to accurately extract complete fault information. To solve this problem, a time–frequency ridge extraction diagnosis method based on multisynchrosqueezing transform (MSST) is proposed. This method utilizes MSST to enhance the compound fault features, especially the gear fault amplitude modulation components. It also utilizes the time–frequency ridge extraction method to separate the gear fault amplitude modulation components and the bearing fault impact pulse components. Additionally, it uses time shifting to substitute data and verifies the independence of the harmonic zero hypothesis to determine the accuracy of the fault components. This method provides a favorable basis for the extraction and identification of compound faults, especially weak faults, in complex dynamic signals of the gearbox. The effectiveness of this method is validated through simulation examples and practical applications.