The Research Wind Farm WiValdi and Advances in Model Predictive Control for Wind Turbines and Farms

Kurzfassung

This is a two-part lecture: In the first part, the Krummendeich research wind farm WiValdi, which went into operation in 2023, is presented. WiValdi currently consists of a meteorological measurement mast, a three-mast array and two multi-megawatt Enercon turbines. A third turbine, rated at 0.5 MW, is currently under construction. Researchers will have access to blade pitch, generator torque, and yaw controls on all three turbines. In the second part of the lecture, the application of real-time Model Predictive Control (MPC) for both wind turbine and wind farm control is discussed, with proof-of-concept validations demonstrated in simulations. For wind turbines, where mechanics and aerodynamics are well understood, a quasi-Linear Parameter Varying (qLPV) model is designed based on first principles and used in a quasi-Linear MPC (qLMPC) scheme. For wind farm control, where airflow between turbines is influenced by wake effects, a data-driven MPC is used. This control algorithm leverages ideas from Koopman theory. The lecture concludes with an outlook on planned research activities in WiValdi.