Real-time Relocation of Floating Offshore Wind Turbines for Power Maximization Using Distributed Economic Model Predictive Control

This paper provides a summary of research on power maximization in floating offshore wind farms. The wind farm control mechanism involves real-time relocation of floating platforms for the purpose of reducing wake overlap along the rotors of downstream machines. Platform displacement is achieved in a passive manner by using aerodynamic forces acting on turbine rotors. The simulation tool and distributed economic model predictive control (DEMPC) scheme used in the investigation are briefly introduced. Additionally, the use of feed-forward neural networks to estimate floating platform dynamics during the optimization process is described. Preliminary simulation results are then presented to demonstrate the effectiveness of the control approach and to gain insight into relevant challenges. For a floating wind farm consisting of three 5 MW turbines aligned with the free stream wind, the DEMPC algorithm yields a 20.2 % increase in energy production relative to traditional greedy operation over the course of a 3,600 sec simulation. The prediction uncertainty of neural networks is also shown to strongly influence controller behaviour.