Wind turbine controllers are nowadays ever more advanced and rely on accurate internal controller model information. Therefore a calibrated model is needed for attaining predictable controller performance and ensuring stable operation. To calibrate the internal model information, a novel learning control scheme has recently been proposed that exploits the dynamics of the closed-loop controlled wind turbine system, without the need for wind speed measurements. The learning algorithm thereby periodically excites the generator power controller input signal. An extremum-seeking demodulation scheme was used to calibrate the internal model information. This paper improves the existing learning scheme in two ways: Firstly, it investigates how the frequency of the excitation signal influences the signal-to-noise ratio. Secondly, the problem was reformulated as a root-finding problem. This requires using the in-phase component of the phase-corrected learning signal. In addition, a precalculated lookup table relates the measured in-phase component directly to model uncertainty. It was found that an increased excitation frequency improves the signal-to-noise ratio (SNR) by an order of magnitude. Combined, these contributions improve the convergence speed more than twenty times, addressing the effect of aerodynamic degradation and its consequences on controller performance.

In wind energy research, airborne wind energy systems are one of the promising energy sources in the near future. They can extract more energy from high altitude wind currents compared to conventional wind turbines. This can be achieved with the aid of aerodynamic lift generated by a wing tethered to the ground. Significant savings in investment costs and overall system mass would be obtained since no tower is required. To solve the problems of wind speed uncertainty and kite deflections throughout the flight, system identification is required. Consequently, the kite governing equations can be accurately described. In this work, a simple model was presented for a tether with a fixed length and compared to another model for parameter estimation. In addition, for the purpose of stabilizing the system, fuzzy control was also applied. The design of the controller was based on the concept of Mamdani. Due to its robustness, fuzzy control can cover a wider range of different wind conditions compared to the classical controller. Finally, system identification was compared to the simple model at various wind speeds, which helps to tune the fuzzy control parameters.

To achieve a high conversion efficiency and at the same time robust control of a pumping kite power system it is crucial to optimize the three-dimensional flight path of the tethered wing. This chapter extends a dynamic system model to account for a realistic, turbulent wind environment and adds a flight path planner using a sequence of attractor points and turn actions. Path coordinates are calculated with explicit geometric formulas. To optimize the power output the path is adapted to the average wind speed and the vertical wind profile, using a small set of parameters. The planner employs a finite state machine with switch conditions that are highly robust towards sensor errors. The results indicate, that the decline of the average power output of pumping kite power systems at high wind speeds can be mitigated. In addition it is shown, that reeling out towards the zenith after flying figure eight flight maneuvers significantly reduces the traction forces during reel-in and thus increases the total efficiency.

Converting the traction power of kites into electricity can be a low cost solution for wind energy. A reliable and robust control system is considered to be crucial for the commercial success of the technology. The focus of this paper is the control of the flight path projected onto the unit sphere. The proposed algorithm is straightforward to implement because it uses mainly LPV and PID control components and is thus easy to certify by the authorities, it allows to define limits for the maximal turn rate to avoid sensor failures, and it allows to use a low gain in the feedback loop to be robust against control loop delays up to 200 ms. This is achieved by splitting the control of the flight path into two different modes of operation: Turn maneuvers and parts of the flight path, where the course angle is constant. During the turning maneuvers mainly feedforward control is used, therefore reducing stability problems. During the straight flight path segments feedback control in combination with Nonlinear Dynamic Inversion (NDI) is used and thus deviations from the planned flight path can be compensated. NDI is needed to compensate the effect of gravity on the turn rate, but also the changes of the steering sensitivity, depending on the apparent wind speed and the angle of attack. A dynamic 4-point model of the kite is used for the validation of the controller performance. The kite is flown in a turbulent 3D wind field using the Mann-model for modeling the turbulence. The results show a low tracking error even in very turbulent wind conditions and even in the presence of large sensor errors and control loop delays: At a turbulence intensity of 26.5% the elevation error was still lower than 1.5°.

Kite power is a promising innovative technology for converting wind energy into electricity at a higher capacity factor and, for many applications, at a lower cost than conventional wind turbines. However, accessing this potential depends substantially on the availability of sophisticated control systems. Delft University of Technology is developing a kite power generator which operates a tethered inflatable membrane wing in a pumping cycle. The flight trajectory is controlled by an actuator unit suspended below the wing and communicates with the ground station control centre via a fast and reliable wireless link. The link is also used to transmit the data of the on-board sensors to the ground. In a future wind park of many kite power systems, the individual kites and ground stations have to communicate among each other, to avoid collisions and to optimize the total energy output of the park. A preparatory analysis has shown that the current prototype would significantly benefit from a distributed control system approach, achieving higher efficiency and increased operational flexibility. For larger installations a distributed control system would be mandatory anyway. For these reasons, a distributed control system with a flexible architecture has been developed. The unique design and first test results are presented.