Application of Wavelet De-noising in Fault Diagnosis of Wind Turbine Gearboxes

摘要

Fault detection and diagnosis of a wind turbine gearbox are important to ensure the reliability and useful life of the wind turbine system. In this paper, an experimental test in wind turbine gearboxes is carried out to obtain faulted information which consists of response random non-stationary noise and the additional response due to failure. An approach, using wavelet de-noising method, is proposed for removing non-stationary noise from the recorded signals. The time domain analysis and frequency analysis are used to diagnose the fault location of the machine accurately. It is shown that wavelet de-noising method provides a better correction than conventional method in order to remove non-stationary noise. Diagnosis results indicate that the high-speed shaft of wind turbine gearboxes has a serious imbalance.