Quasi Linear Parameter Varying modeling for wind farm control using the 2D Navier-Stokes equations

Controlling the flow within wind farms, to increase the overall energy capture and to reduce the fatigue loads, is a challenging control problem due to its size and the underlying time-varying dynamics. State-of-the-art wind farm controllers follow the open-loop control paradigm by computing optimal set-points using static models. However, due to the time-varying inflow conditions and uncertain flow dynamics the closed-loop control paradigm should be used. In this work, a control-oriented quasi Linear Parameter Varying (LPV) descriptor model will be derived using the two dimensional (2D) Navier-Stokes equations. Sparsity of the system matrices make it possible to use this model for closed-loop controller synthesis. Some key features of the approach will be presented using a two turbine test case.