A Preliminary Demonstration of a Control Co-design Framework for the Floating Offshore Wind Turbines

The traditional sequential method is frequently used for the design of the complex floating offshore wind turbines, which takes each sub-system sequentially and the control optimization is always the last step for manipulating a specific wind turbine. This may simplify the problem by focusing on a single subsystem in each step, but it may miss the global optimum due to neglecting the couplings between different sub-systems. The control co-design is a solution by integrating all sub-systems in a single formula and solving this formula for the global optimum. In this study, a control co-design framework is developed for the floating offshore wind turbines. In this framework, the blade airfoil chord and twist distributions can be parameterized with the other structural properties (mass, stiffness, bending modes, etc.) required by the simulation updated automatically. Based on the values of pre-defined control variables, the variable-speed and blade pitch controllers can be re-tuned automatically considering the dynamics of the new system. The optimizer finds the next iterations of the structural and control variables until termination. Using a reference floating wind turbine model, which is generated by the sequential method, the annual energy production is maximized for better power generation capability while the blade mass is minimized simultaneously for lightweight and cost-reduction purposes. The constraints of design variables, fatigue loads, ultimate loads, and blade deflection are defined for the optimization problem. To solve this high-dimensional, nonlinear, and highly constrained problem, the genetic algorithm is used to find a Pareto front for the trade-off between exploration and exploitation. The design alternatives in the Pareto front show a great reduction in blade mass and a slight increase in annual energy production. This demonstrates the improvement made by the control co-design method.