Final Report Summary - HIGHWIND (Simulation, Optimization and Control of High-AltitudeWind Power Generators)
Airborne Wind Energy (AWE) is a young wind power technology that forms an alternative to classical wind turbines. Instead of a rotor on top of a heavy and expensive tower, AWE is based on autonomously flying wings that are connected to the ground by a thin but strong tether. Power is generated by periodically reeling in and out the tether, subject to tether tension variations due to an intelligently chosen flight cycle. AWE systems use less material per power, and can reach higher altitudes and therefore tap into stronger winds than classical wind turbines. Also, due to their small mass footprint, they might be easier to deploy off-shore and in remote locations. The focus of the HIGHWIND project was on the optimization, control and estimation problems inherent to AWE systems, and the HIGHWIND team has made the following major contributions.
First, AWE systems are very unstable systems, and have very intricate flight characteristics. The numerical optimization of their performance in terms of power-generation or other objectives is a very involved problem that was addressed by the HIGHWIND team, which developed open-source software and systematic strategies to tackle their optimization efficiently and reliably. The software package CasADi was significantly refined based on the experience of the team, and was made available as open-source code to the AWE community as well as all other advanced engineering technology developers.
Second, the HIGHWIND project has realised a new experimental demonstrator, the outdoors carousel. It consists of a rotating arm, a control computer, a winch, a tether with variable length, and an electronically equipped small airplane attached at its end. This setup was used to test a variety of setups and control concepts, and is still in operation. It shall be used to demonstrate the possibility to start up AWE systems with a rotary motion and transition into power generating orbits. The embedded control hardware developed and used during the HIGHWIND project was made available to the public as open hardware.
Third, the HIGHWIND team has developed advanced optimization based control and estimation strategies and tested them experimentally on the experimental setup. These strategies allow for controlling and maximizing the performance of the controlled systems while respecting their intricate limitations, and allow for efficiently combining the information provided by the numerous sensors. The numerical methods and underling software tools have been extensively studied in simulations and experiments. The resulting open-source software tools, in particular ACADO Code Generation, are increasingly used in many industrial fields such as mechatronics, robotics, and the automotive industry.
First, AWE systems are very unstable systems, and have very intricate flight characteristics. The numerical optimization of their performance in terms of power-generation or other objectives is a very involved problem that was addressed by the HIGHWIND team, which developed open-source software and systematic strategies to tackle their optimization efficiently and reliably. The software package CasADi was significantly refined based on the experience of the team, and was made available as open-source code to the AWE community as well as all other advanced engineering technology developers.
Second, the HIGHWIND project has realised a new experimental demonstrator, the outdoors carousel. It consists of a rotating arm, a control computer, a winch, a tether with variable length, and an electronically equipped small airplane attached at its end. This setup was used to test a variety of setups and control concepts, and is still in operation. It shall be used to demonstrate the possibility to start up AWE systems with a rotary motion and transition into power generating orbits. The embedded control hardware developed and used during the HIGHWIND project was made available to the public as open hardware.
Third, the HIGHWIND team has developed advanced optimization based control and estimation strategies and tested them experimentally on the experimental setup. These strategies allow for controlling and maximizing the performance of the controlled systems while respecting their intricate limitations, and allow for efficiently combining the information provided by the numerous sensors. The numerical methods and underling software tools have been extensively studied in simulations and experiments. The resulting open-source software tools, in particular ACADO Code Generation, are increasingly used in many industrial fields such as mechatronics, robotics, and the automotive industry.