Open Access

The transition of our energy system towards a generation by renewables, and the corresponding developments of wind power technology enlarge the requirements that must be met by a wind turbine control scheme. Within this thesis, the role of modern, model-based control approaches in providing an answer to present and future challenges faced by wind energy conversion systems is discussed. While many different control loops shape the power system in general, and the energy conversion process from the wind to the electrical grid specifically, this work addresses the problem of power output regulation of an individual turbine. To this end, the considered control task focuses on the operation of the turbine on the nonlinear power conversion curve, which is dictated by the aerodynamic interaction of the wind turbine structure and the current inflow. To enable a power tracking functionality, and thereby account for requirements of the electrical grid instead of operating the turbine at maximum efficiency constantly, an extended operational range is explicitly considered in the implemented control scheme. This allows for an adjustment of the produced power depending on the current state of the electrical grid and is one component in constructing a reliable and stable power system based on renewable generation. To account for the nonlinear dynamics involved, a linear matrix inequalities approach to control based on Takagi-Sugeno modeling is investigated. This structure is capable of integrating several degrees of freedom into an automated control design, where, additionally to stability, performance constraints are integrated into the design to account for the sensitive dynamical behavior of turbines in operation and the loading experienced by the turbine components. For this purpose, a disturbance observer is designed that provides an estimate of the current effective wind speed from the evolution of the measurements. This information is used to adjust the control scheme to the varying operating points and dynamics. Using this controller, a detailed simulation study is performed that illustrates the experienced loading of the turbine structure due to a dynamic variation of the power output. It is found that a dedicated controller allows wind turbines to provide such functionality. Additionally to the conducted simulations, the control scheme is validated experimentally. For this purpose, a fully controllable wind turbine is operated in a wind tunnel setup that is capable of generating reproducible wind conditions, including turbulence, in a wide operational range. This allows for an assessment of the power tracking performance enforced by the controller and analysis of the wind speed estimation error with the uncertainties present in the physical application. The controller showed to operate the turbine smoothly in all considered operating scenarios, while the implementation in the real-time environment revealed no limitations in the application of the approach within the experiments. Hence, the high flexibility in adjusting the turbine operating trajectories and structural design characteristics within the model-based design allows for efficient controller synthesis for wind turbines with increasing functionality and complexity.