Model Predictive wind turbine control for load alleviation and power leveling in Extreme Operation conditions

Modern wind turbine control systems do not only have to guarantee safe and efficient operation, but also reduce mechanical loads to a minimum and by that allow for lighter weight turbine designs. In order to limit the amount of required control loops and corresponding tuning parameters, more and more attention is paid to multivariable feedback control which handle all aspects in only one framework. This paper presents a Model Predictive Controller (MPC) designed for a commercial 3MW wind turbine. The goal of that controller is to level output power while mechanical loads are limited. Although the reduced order model required for the MPC and presented in this paper, is of minimal order, it is sufficiently detailed for the application at hand. The MPC is compared with the controller currently installed. Simulation based tests with a fully validated multibody model show that the MPC can maintain the output power during an Extreme Operation Gust (EOG) within a limit of only 4%, compared to 8% before, and at the same time can reduce the tower base moment by approximately 15%.