Wind turbine yaw control plays an important role in increasing the wind turbine production\nand also in protecting the wind turbine. Accurate measurement of yaw angle is the basis of an effective\nwind turbine yaw controller. The accuracy of yaw angle measurement is affected significantly by\nthe problem of zero-point shifting. Hence, it is essential to evaluate the zero-point shifting error\non wind turbines on-line in order to improve the reliability of yaw angle measurement in real time.\nParticularly, qualitative evaluation of the zero-point shifting error could be useful for wind farm\noperators to realize prompt and cost-effective maintenance on yaw angle sensors. In the aim of\nqualitatively evaluating the zero-point shifting error, the yaw angle sensor zero-point shifting fault is\nfirstly defined in this paper. A data-driven method is then proposed to detect the zero-point shifting\nfault based on Supervisory Control and Data Acquisition (SCADA) data. The zero-point shifting\nfault is detected in the proposed method by analyzing the power performance under different yaw\nangles. The SCADA data are partitioned into different bins according to both wind speed and yaw\nangle in order to deeply evaluate the power performance. An indicator is proposed in this method\nfor power performance evaluation under each yaw angle. The yaw angle with the largest indicator\nis considered as the yaw angle measurement error in our work. A zero-point shifting fault would\ntrigger an alarm if the error is larger than a predefined threshold. Case studies from several actual\nwind farms proved the effectiveness of the proposed method in detecting zero-point shifting fault\nand also in improving the wind turbine performance. Results of the proposed method could be\nuseful for wind farm operators to realize prompt adjustment if there exists a large error of yaw\nangle measurement.
Loading....