8th International Conference on Advanced Power System Automation and Protection (APAP2019)

898 / 2019-04-30 15:14:59

Fault Detection Method for Wind Turbine Gearbox Based on Sparse Bayesian Probability Estimation

Fault detection, Probability interval, SCADA data, Spare Bayesian Learning, Wind turbine.

In this paper, a SCADA data-based fault detection method for gearbox oil filtration pressure is proposed, and the core of the method is Spare Bayesian Learning (SBL) algorithm. According to training the historical normal operation data from SCADA system, the gearbox oil filtration pressure estimation model based on SBL can be constructed. Based on the model, the probability distribution interval of the gearbox oil filtration pressure can be estimated. Then, the abnormal state of gearbox oil filtration pressure can be judged by observing whether the actual data within the probability distribution interval. In addition, statistical hypothesis testing method is used to verify the reliability of the anomaly detection results. Through the method, the oil filtration pressure abnormal detection problem can be transformed into a parameter estimation problem with low computational complexity. Case studies are conducted on two known fault WTs, and the results demonstrate the effectiveness of the method.