It is proposed to estimate wind velocity, Angle-Of-Attack (AOA) and Sideslip Angle (SSA) of a fixed-wing Unmanned Aerial Vehicle (UAV) using only kinematic relationships with a Kalman Filter (KF), avoiding the need to know aerodynamic models or other aircraft parameters. Assuming that measurements of airspeed and attitude of an UAV are available as inputs, a linear 4th order time-varying model of the UAV's longitudinal speed and the 3-D wind velocity is used to design a Kalman-filter driven by a GNSS velocity measurement airspeed sensor. An observability analysis shows that the states can be estimated along with an airspeed sensor calibration factor provided that the flight maneuvers are persistently exciting, i.e. the aircraft changes attitude. The theoretical analysis of the KF shows that global exponential stability of the estimation error is achieved under these conditions. The method is tested using experimental data from three different UAVs, using their legacy autopilot to provide basic estimates of UAV velocity and attitude. The results show that convergent estimates are achieved with typical flight patterns indicating that excitation resulting from the environment and normal flight operation is sufficient. Wind velocity estimates correlate well with observed winds at the ground. The validation of AOA and SSA estimates is preliminary, but indicate some degree of correlation between the AOA estimate and vertical accelerometer measurements, as would be expected since lift force can be modeled as a linear function of AOA in normal flight

Abstract-Structural changes due to ice accretion are common causes for unmanned aerial vehicle incidents in Arctic regions. For fixed wing unmanned aerial vehicles (UAVs) the leading edge of airfoil surfaces is one of the primary surfaces exposed to these changes, causing a significant reduction in aerodynamic ability, i.e. decreasing lift and manoeuvrability, and increasing drag, weight, and consequently power consumption. Managing or altogether preventing ice accretion could potentially prevent icing related UAV incidents and increase the operability of UAVs.This paper addresses the issue of structural change, caused by ice accretion, on small UAVs by integrating a power control system and an electrically conductive carbon nano material based coating for temperature control of UAV airfoil surfaces. Performance assessment is achieved through extensive laboratory experiments, where various coating layouts have been investigated in various conditions, with temperatures ranging from +25 • to -25 • . The experimental setup consists of an Arduino microcontroller capable of controlling power delivery to the coating through feedback from thermocouples and a humidity sensor, sensing the surface temperature of the leading edge of the UAV wing and ambient humidity, respectively. Experiments reveal that a layout, where the coating covers the entire length of an wing is preferable, with the solution being highly capable of rapidly increasing the airfoil surface temperature (de-icing) when needed, and of maintaining an approximately constant airfoil surface temperature (anti-icing) when needed, all the while keeping power and energy consumption within weight and cost constraints imposed by the small scale of the UAV. The results represents a proof of concept by using an electrically conductive coating for de-icing and anti-icing of leading edge UAV airfoils.

No abstract

UAV icing is a severe challenge that has only recently shifted into the focus of research. Today, there are no mature icing mitigation technologies for UAVs, except for the largest fixed-wing drones. We are working on the development of an electro-thermal icing protection technology called D•ICE for medium-sized fixed-wing UAVs. As part of the design process, an experimental test campaign at the Cranfield icing wind tunnel has been conducted. This paper describes the icing protection system and shares experimental results on its capability for icing detection and anti-icing. Icing detection is based on an algorithm evaluating temperature signals that are induced on the leading-edge of the wing. A baseline signal is generated during dry (icing cloud off) conditions and compared to a signal during wet (icing cloud on) conditions. Due to significant differences in the heat transfer regime, the system can differentiate between these two states. The experiments show that our system can reliably detect icing conditions based on this principle. Furthermore, the anti-icing capability of the system is proven for two icing cases. The minimal required heat flux to keep the surface ice-free was obtained by gradually reducing power supply to the heating zones until icing could be detected. These experimental results were compared to FENSAP-ICE simulations. The test campaign includes a successful fully-autonomous run, where the system automatically detected icing and initiated suitable anti-icing measures.

Abstract-This paper tackles the problem of power regulation for wind turbines operating in the top region by an adaptive passivity based individual pitch control strategy. An adaptive nonlinear controller that ensures passivity of the mapping aerodynamic torque-regulation error is proposed, where the inclusion of gradient based adaptation laws allows for the on-line compensation of variations in the aerodynamic torque. The closed-loop equilibrium point of the regulation error dynamics is shown to be UGAS (uniformly globally asymptotically stable). Numerical simulations show that the proposed control strategy succeeds in regulating the power output of the wind turbine despite fluctuations of the wind field due to wake and turbulence, without overloading the pitch actuators.