An economic NMPC formulation for wind turbine control

Model Predictive Control (MPC) is a strong candidate for the control of large Multi-MegaWatt Wind Turbine Generators. Several MPC and some Nonlinear MPC scheme have been proposed in the literature, based on reference-tracking objective functions. While the resulting schemes offer very promising results, the difficulty of tuning a reference-tracking NMPC scheme for performance is likely to be a hindrance to the industrial success of NMPC-based WTG control. Because they directly maximize the system performance, economic NMPC schemes are more straightforward to tune. Economic NMPC schemes present, however, some known difficulties that are a serious obstacle to their real-time deployment. This paper presents an economic NMPC formulation for maximizing the generated power of wind turbine generators, which does not suffer from such difficulties. The relationship between the proposed and more classical reference-tracking approaches is formally established.