Hierarchical wind farm control for power/load optimization

The paper describes the hierarchical control concept for wind farm power/load optimization with emphasis on compensation of disturbances. The control concept was developed for the EU-FP7 project Distributed Control of Large-Scale Offshore Wind Farms – Aeolus. The Aeolus project approaches the wind farm control problem by considering wind turbines in a wind farm as individual power actuators affected by constraints. It is assumed that the demanded wind farm power production is given in advance, i.e., the power reference value for the whole farm is provided by the Transmission System Operator (TSO). The supervisory farm controller then distributes the power demand among wind turbines based on the dynamic model of wind field inside the wind farm. The objective of the distribution is to minimize the cumulative load of all wind turbines in the wind farm, while respecting the system constraints and fulfilling the TSO demands. The presented control concept consists of two levels: i) the high level that acts on a slower time scale and computes the optimal distribution of production and loads, and ii) the low level of control that acts on the faster time scale and ensures the tracking of the optimal distribution under disturbances. In this paper the focus is on the control design for the low level of control. To minimize the computational effort the control design procedure replaces the online optimization with the off-line optimization approach based on the principles of multi-parametric programming.