Conventional GPS and encoder-based solutions for the real-time position and velocity tracking of tethered aircraft in airborne wind energy (AWE) systems are known to have some practical limitations. As an alternative, the main contribution of this paper is to present a setup based on 2.4 GHz radio-frequency (RF) devices, for which there are still no reported experimental results in the literature. To this end, four estimation algorithms are formulated to a lateration problem: linear and nonlinear least squares (kinematic approaches), as well as the extended and the unscented Kalman filters (dynamic approaches). The Cramér–Rao lower bound is computed to serve as a benchmark for the filtering performance in terms of the position estimate covariance. Drawbacks of the two kinematic algorithms are discussed, followed by the advantages of using the two Kalman filters in a sensor fusion scheme that uses not only the RF rangings but also line angles and length measurements obtained from rotary encoders at the ground station. Estimates computed from field experimental data acquired with a small-scale AWE prototype are validated against the encoder measurements. The results show that setups based on RF ranging devices have many advantages over more conventional positioning techniques such as those based on line-following mechanisms with encoders, GPS and IMU, and computer vision, representing a promising solution for AWE applications.

A new filtering approach is presented, able to estimate online the aerodynamic characteristics of tethered wings used in airborne wind energy systems. The approach is based on the extended Kalman filter (EKF) and on a representative model of the flight dynamics. The filter is fed with measurements available at the ground station, namely, the line angles and their rates, the traction force on the tether, and the wind speed and direction a few meters above the ground. Convergence and accuracy of the estimates are first evaluated through simulations. Then, the approach is demonstrated using extensive experimental data, collected during field tests with two small-scale prototypes built independently by different research groups and four different flexible airfoils (power kites), of both ram-air type and of leading-edge-inflated type. The experimental results show that the algorithm can effectively estimate both the state and the aerodynamic parameters of the wings, as well as the wind speed and direction at the wing's altitude, and makes it possible to analyze the kite aerodynamics during highly dynamical tethered flight.

Magnus effect-based Airborne wind energy (AWE) systems are a promising yet still unexplored concept for harnessing wind power at high-altitudes. While other aspects of the technology have been recently studied, the problem of obtaining accurate information regarding the aerodynamic behavior of the suspended cylinder as its spin ratio varies remains open. This paper presents an adaptation of an existing estimation strategy based on a constrained Extended Kalman filter (EKF) for the aerodynamic characterization of a small-scale Magnus effect-based AWE prototype. The evaluation is performed on data obtained during wind tunnel experiments, and results indicate that, after minor modifications, the chosen approach can indeed be applied to Magnus effect-based AWE systems. Moreover, provided that the cylinder's angular velocity is available, it can be employed for approximately determining the relationship existing between the aerodynamic coefficients of lift and drag and the spin ratio of the airborne structure.

This paper presents an estimation scheme for Airborne Wind Energy systems based on the Extended Kalman Filter and on a simple, yet representative, dynamic model of the wing flight. Besides estimating the airfoil position, velocity, and acceleration, the filter also estimates the wind vector and the aerodynamic forces acting on the wing. Even though our setup currently fuses ground based measurements only, it could be easily extended to employ measurements from airborne sources such as GPS receivers, ranging devices, accelerometers and airspeed sensors. Moreover, our approach allows for the straightforward computation of other important variables such as the angle of attack and the equivalent aerodynamic efficiency. An established simulation model is used to validate the filter, which is later evaluated on experimental data from a field test. Finally, we discuss the possibility of applying the proposed approach in the context of system monitoring, control, and optimization.

This paper presents a 2D model describing the flight dynamics of a power kite on the tangent plane of its wind window. This model depends on variables that can be easily measured, as well as on parameters that can be readily identified. To this end, an Extended Kalman filter is employed on-line, allowing us to cope with instrumentation noise while identifying the steering gain and transport delay. Taking advantage of the plant knowledge, we design a course angle controller to ensure closed-loop stability. The proposed filtering and control scheme is validated based on a hardware-in-the-loop setup. Results of field tests are also presented.