Wind turbine design is an interesting field for research due to the complexity of the system and its multi-disciplinary nature. Historically, the design of wind turbines has been done sequentially, where the aerodynamics, the structure, and the controller are designed one after the other. However, this approach does not allow us to take advantage of the potential couplings between the different disciplines. Instead, designing simultaneously the different aspects of the turbine can lead to significant improvements in terms of power capture or cost savings. Among the different components of the turbine, the controller is of particular interest to be more closely integrated into the design process, due to its role in load alleviation, stability, and power production.

However, there are important challenges to tackle in order to take a multi-disciplinary approach to the design process. The literature in this field has focused on leveraging existing design tools and numerical methods for this purpose. In particular, numerical optimization has shown to be a powerful tool for supporting the design process due to its ability to solve large and complex problems.

The research work presented in this thesis explores how to include the controller design in the design process of the wind turbine, in the context of numerical optimization. This approach, called control co-design, has emerged from the broader field of dynamic systems design and has shown promising results when applied to wind turbine design. This thesis contributes to this research area by quantifying the benefits of control co-design using state-of-the-art analysis models and design problem formulations, from the physical design perspective. The emphasis is placed on computational effort and fair comparison with existing design processes.

In this thesis, we apply control co-design to several design problems and present numerical methods to facilitate the use of control co-design with efficient use of computational power. Our results show that the control co-design approach is not beneficial in all cases. There are three main research contributions, covering both the operational and dynamic aspects of controller design. First, we develop a method to include the design of the power regulation strategy in wind turbine optimization to align control and physical design objectives, resulting in improved steady-state performances. Second, we apply design sensitivity analysis techniques to estimate the benefits of control co-design for designs driven by load constraints. Our results show that a control co-design approach presents little benefit in the presence of an active frequency constraint that is not impacted by a change in control design. Finally, we compare different design processes to include openloop dynamic control design in rotor design optimization, showing that a simultaneous control co-design approach does not outperform a sequential or iterative-sequential design process.