Wake losses in wind farms, caused by closely spaced turbines, reduce downstream power production and overall efficiency. These losses can be mitigated through wind farm flow control strategies, specifically wake steering, which involves yaw control to misalign turbines with incoming wind, and redirecting wakes to enhance overall power output. This thesis aims to develop a real-time, model-based controller that addresses operational and environmental challenges in large-scale wind farms. The controller adapts to off-design conditions such as changing atmospheric factors and offline turbines, making it more effective than traditional controllers based on pre-optimized lookup tables. The research modifies engineering wake models to account for heterogeneous wind conditions, improves the computational efficiency of the Serial-Refine optimization strategy for real-time use, and incorporates tuning of wake model parameters to ensure adaptability. Additionally, the impact of offline turbines on optimal yaw misalignment angles is examined. Results indicate that adapting wake models using sensor data improves power predictions in non-uniform wind environments, while the distributed optimization framework for real-time yaw adjustments performs comparably to centralized methods. Incorporating offline turbines into the control strategy further boosts power output, leading to increased revenue for the wind farm. This research advances wind farm flow control by developing control policies that are tailored to dynamic, real-world site conditions.

Floating offshore wind turbines (FOWTs) unlock the potential to harness energy from wind in deeper waters. Despite their potential, the major obstacle to large-scale commercial deployment remains the floating substructure's high cost. Multidisciplinary design, analysis and optimisation techniques are commonly employed to improve their cost-competitiveness, but existing models often use simplified engineering models. This approach risks neglecting design considerations typical of structures subjected to complex aero-hydro-servo-elastic loads.

To address the need for incorporating higher-fidelity analysis methods, an optimisation framework is developed, integrating OpenFAST to simulate the response of the FOWT system under various environmental and operational conditions. Leveraging the flexibility of the Python framework and the wealth of output data contained within the OpenFAST simulation results, this holistic approach offers opportunities to explore novel and cost-effective platform designs with high reliability.

To demonstrate its effectiveness, the optimisation framework is utilised to achieve a significant reduction of the DeepCWind platform's structural mass. Among the considered constraints, the platform pitch motion was critical, reaching maxima for design load cases characterised by the most extreme wind and waves. Although the inclusion of a structural model for the platform comes at considerable computational expense, it enhances the framework's value, as structural integrity can be verified.

The recommended use of the optimisation framework is for in-depth studies, following preliminary design explorations conducted with cheaper models. Future efforts should focus on extending the structural and hydrodynamic model to improve the framework's versatility, and make it applicable to a wider range of platform concepts and turbine sizes.

The growing importance of renewable energy to meet the demands of growing population has driven much focus for research in the wind energy sector. Currently most of the power production in the wind energy sector is done using the Horizontal Axis Wind Turbine (HAWT). due to its higher efficiency and reliability as compared to Vertical Axis Wind Turbine (VAWT). However, due to limitations of HAWTs such as the complex yaw mechanism, higher positioning of the center of mass and relatively difficult up-scaling, VAWT are receiving more attention.

Attempt to access high range of power from VAWTs at offshore locations may damage the turbine blades due to increased loads. This thesis is dedicated to reduce the periodic disturbances on the turbine blades of VAWTs without affecting the total power production in a rotation, thereby ensuring the reliability and safe operation of VAWT. A control technique called Subspace Predictive Repetitive Controller (SPRC) is used for the recursive identification to estimate the parameters of wind turbine model and further providing an optimal control law accordingly. Basis functions have been used to reduce the dimensionality of the system, following which the system identification has been performed in the lifted domain.

Using the identified model, two types of control approaches have been applied. In the first approach, the objective function is to track a specific reference that indicates blade load. This reference ensures a reduction in peak loads of the VAWT by transferring the loads from their upstream to their downstream part, without comprising on the power production in one whole rotation of VAWT. In the second approach, a general control strategy is used in which the controller is given freedom to choose the pitch trajectory so as to reduce the blade loads. The controller is specified with a power reference that ensures the maintenance of total power. This strategy allows the turbine to be operated in different wind conditions, as a higher weighting is assigned to power production when wind speeds are less than rated wind speeds, and a higher weighting is assigned to the reduction of blade load when wind speeds are more than the rated wind speeds. .

These results show a real potential of the data-driven SPRC approach in wind turbines, and highlights new mechanisms to reduce the turbine loads on \ac{VAWT}s. These mechanisms offer valuable insights into enhancing the functions of VAWTs in a more reliable and damage-free manner.