Altitude Control for Optimal Performance of Tethered Wind Energy Systems

Tethered wind energy systems replace conventional rigid towers with flexible cables, and can harness strong high-altitude winds using as little as 10 percent of the material required by traditional turbines. Levelized costs for tethered systems are estimated to be from $0.05 to $0.25 per kW-h, providing a cost-competitive energy solution for remote off-grid communities, military bases, and deep-water offshore locations. Tethered systems can easily change altitude, allowing them to achieve optimal performance by finding the best available wind velocity. Preliminary research suggests that exploiting this additional operational freedom can increase energy production by as much as 50 percent over fixed-altitude tethered turbines. This award will develop control laws that vary tether length to measure the change in wind speed with altitude, and use that information to increase energy production. University students working on this project will gain real-life practical experience through collaboration with Altaeros Energies, a tethered wind energy startup based in Boston, Massachusetts. Outreach activities with high school students in North Carolina will introduce the role of STEM disciplines in developing new renewable energy resources, and broaden participation by non-traditional and underrepresented groups.

The ability to optimize the altitude of a tethered wind energy system depends both on knowledge of the wind shear profile and robust optimization strategies that maximize net energy generation. The research will attack this dual problem of wind shear profile mapping and altitude optimization through the fusion of information maximization and extremum seeking control techniques. While both tools are powerful in their own right, neither provides an ideal stand-alone framework for the simultaneous characterization of wind shear profile and optimization of altitude. Two mechanisms for fusing the mapping and optimization objectives will be pursued in this research: one is an implicit mechanism, using stochastic receding horizon control, and the other is an explicit method that estimates the map's entropy to morph the perturbation signal for the extremum seeking algorithm. These methods will be validated using historical wind profiles from NOAA and NASA and will be flight tested on the Altaeros Energies Buoyant Airborne Turbine (BAT). The tools arising from this research will not only benefit tethered wind energy organizations but will also improve fundamental understanding of performance optimization in poorly modeled and time-varying environments.