Note: This content is accessible to all versions of every browser. However, this browser does not seem to support current Web standards, preventing the display of our site's design details.


Automatic Power Cycles for Airborne Wind Energy Generators


A. Zgraggen


Airborne wind energy (AWE) is an emerging technology to harvest the energy of wind at higher altitudes than traditional wind turbines and thus improving important shortcomings of such turbines. One of the most promising concepts in the field of AWE are the ones using crosswind motion where a wing tethered to the ground flies roughly perpendicular to the wind flow and thus exerting large aerodynamic forces. This idea emerged around forty years ago but was never set into practice due to prohibitive engineering problems until in the last decade when various groups picked up this idea again. Two basic concepts are mainly investigated, the drag and lift mode. The first flies at constant tether length and uses on-board generators and feeds the power down to the ground. The latter directly uses the traction force on the tether to drive a generator on the ground by unreeling the tether. Due to the reeling during power generation a two phase cycle has to be flown. In the first phase, the traction phase, power is produced and in the second phase, the retraction phase, the tether is retracted using only a fraction of the previously produced power hence leading to a positive energy balance. In this thesis we will focus on the lift mode. However, many of the results can also be directly applied to the drag mode.

Automatic control is a key aspect for these kind of systems since they aim to run continuously for many hours under varying environmental conditions %while still respecting operational constraints. This control problem involves fast, non-linear, unstable time-varying dynamics subject to hard operational constraints and external disturbances. The control problem of the traction phase aims at making the wing flying along a crosswind path which yields the highest traction force. This problem was investigated by various research groups and companies however only little progress has been made on actual implementations on real systems due to the stringent assumptions made during the controller synthesis and on available information. On the other hand, the control problem of the retraction phase has been widely neglected and only few theoretical results are available to the public.

In this thesis we aim to fill this gap between theoretical results and actual implementation on a real system. We propose flight controllers for the traction and retraction phase of a ground-based generation AWE system, i.e. lift mode, which are straightforward to implement and tune and do not need information which are difficult to obtain and only employ directly measurable quantities. In contrast with most existing approaches, the controllers do not require an estimate of the wind situation or a measure of the apparent wind at the wing's location but only a rough estimate of the wind direction to start the system which is readily available by ground based anemometers.

The traction phase controller is based on the notion of the velocity angle which is a variable representing the wing's flying direction based on its velocity vector and is therefore particularly suited for feedback control. In addition, the proposed approach features few parameters whose effects on the system's behavior are very intuitive, hence simplifying the tuning procedure.

For the retraction controller two different approaches are investigated. A first approach is based on the traction controller also using the notion of the velocity angle such that only the high-level guidance strategy has to be changed. However, due to the low relative speed of the wing during the retraction phase the definition of the velocity angle has to be adapted and requires an estimate of the wind direction at the wing's location. Since such an estimate is not straightforward to compute, an alternative approach using only directly measurable quantities is derived, hence resulting in a more robust and reliable solution.

The proposed control systems are tested in simulation and experimentally on two different small-scale prototypes, with varying wind conditions and using different wings. Only minimal information of the wind situation are needed by the employed control system, in particular just a rough estimate of the wind direction. However, the wind field changes over time and distance, especially in the vertical axis known as the wind shear effect. In the last part of the thesis we therefore propose a real-time optimization and adaptation algorithm to maximize the power output by optimally positioning the flown crosswind path and adapting this position in real-time to changing wind conditions without directly measuring the wind but only using measurements of the traction force on the tether. The algorithm is able to adapt the control system's parameters in real-time to maximize the average traction force with uncertain and time-varying wind. The presented algorithm is not dependent on a specific hardware setup and can act as an extension of existing control structures.

Download available here


Type of Publication:

(03)Ph.D. Thesis

M. Morari

No Files for download available.
% Autogenerated BibTeX entry
@PhDThesis { Xxx:2014:IFA_5006,
    author={A. Zgraggen},
    title={{Automatic Power Cycles for Airborne Wind Energy Generators}},
    school={ETH Zurich},
Permanent link