Modelling and Control of a Magnus Effect-Based Airborne Wind Energy System

Airborne wind energy systems (AWE systems) are systems that use the energy of the wind at altitudes beyond the reach of conventional wind turbines, or at locations where erecting a conventional wind turbine is prohibitively complex, by using an airborne device (aircraft) instead of turbine blades and ropes (tethers) instead of a turbine-carrying tower. Compared with conventional wind turbines, AWE systems provide access to stronger and steadier high-altitude winds to realize significantly larger energy production and facilitate a smaller ratio of construction material per unit of produced energy. Achieving their required autonomous operation necessitates a well-designed control system, which is usually based on a mathematical model of system dynamics. This thesis considers an AWE system using the Magnus effect to induce aerodynamic forces driving a ground-based generator through a rope. The thesis first develops a system dynamics model, including the specific components such as the airborne device, the rope connecting the airborne module with the ground-based winch, and the winch connected to the generator. The emphasis is on the airborne module motion in the vertical plane aligned with the wind direction, making the planar motion dynamics model the main modelling contribution. Models of different complexities with respect to tether modelling accuracy are developed and compared regarding their utilization in subsequent control system development. The resulting model is then applied in simulation and optimization-based analyses, where an optimization problem is formulated to optimize the two control variables, the speed of rotation of the airborne module cylinder and the speed or force of the unwinding tether, to maximize energy produced during a continuously repeatable operating cycle. Next, the obtained optimal behaviour of the system is physically interpreted to provide the physical foundation required for control system synthesis. The physical insights obtained in this manner are then used to design a control system yielding energy production close to optimal. The emphasis is on supervisory control strategy accounting for dynamics of low-level cylinder motor and generator control systems. Finally, the developed control strategy is assessed and analysed in simulations regarding energy production quality and robustness against main operational parameters. Based on the corresponding comparative analysis, the final recommendations for implementing the control strategy are provided.