Incorporating environmental impacts into zero-point shifting diagnosis of wind turbines yaw angle

Abstract

The diagnostic mechanism of yaw angle zero-point shifting (ZPS) of wind turbines (WTs) is based on the variation of the output power at different yaw angles. However, the output power is significantly affected by environmental factors. To improve the diagnostic performance, a diagnostic method based on SCiForest and Sparse Gaussian Process Regression (SGPR) is proposed, in which two environmental factors including air temperature and turbulence intensity are incorporated into the ZPS diagnosis procedure. On this basis, the diagnostic model is developed, the diagnostic performance is explored by considering the individual and coupling effect of environmental factors, and the model validation is made on the actual operation data of the three 3 MW WTs. The validation results show that the air temperature and turbulence intensity have a mutual promotion effect on the diagnostic performance. Specifically, the diagnostic accuracy for the three WTs is slightly improved by including the individual effect of the environmental factors, while it is noticeably enhanced by 68.706 %, 65.411 % and 59.572 %, respectively, considering the coupling effect. After calibrating the yaw angle ZPS, the annual profit for the three WTs can be increased by 9.150 %, 15.417 % and 21.649 %, respectively, which shows the proposed method has the potential in improving the operation efficiency of the WTs and reducing the cost of wind power.