As a renewable source of energy, wind is widely used to produce electrical power. The progress of wind turbine technology can greatly benefit from the improvement of control algorithms. The pitch angle con- trol of a horizontal axis wind turbine above the rated wind speed is a challenging issue related to the nonlinear aerodynamic behavior of blades. The linearization of aerodynamic model around nominal op- erating condition, as well as manufacturing deficiencies, result in unknown parameter uncertainties in a wind turbine model. Therefore, the performance of controller, which is designed based on the mathemat- ical model, defects in practice. In the current paper, an adaptive self-tuning regulator (STR) configuration is proposed for the pitch control, so that the parameters of wind turbine model are constantly estimated and the controller gains are updated based on the assessed parameters. The STR structure consists of a recursive least square estimator and a proportional-integral-derivative (PID) controller with adjustable gains, which are determined by the pole placement method in a real-time routine. The robustness of the closed loop system is investigated by implementation of the control structure on an aero-servo-elastic wind turbine simulator. For the sake of comparison, a baseline gain scheduling PID controller, which is well-accepted for wind turbine pitch control, is designed. A comparison between the simulations of two controllers confirms a significant improvement in the closed-loop performance including less fluctuation of rotor speed and power besides minor fatigue loads on the blades and main-shaft. ©2016ElsevierLtd.

Keywords