Control system development for autonomous soaring

Thermal and dynamic soaring are two techniques commonly used by birds to extract energy from the atmosphere. This enables them to reduce, energy used during flight and increases their endurance. The thermal soaring technique involves extraction of energy from thermal updrafts and in dynamic soaring energy is extracted from wind shear. These techniques are investigated in this thesis using point mass and non-linear 6DoF models of an unmanned powered sailplane. The key challenges of autonomous thermal soaring are the ability to identify remote thermal activity using on-board sensors and to position correctly in a thermal. In dynamic soaring, a real-time fuel saving trajectory generation technique along with a trajectory following control system is needed. A hand held IR camera was used to assess the feasibility to observe hot spots associated with thermals. The thermal positioning capability was demonstrated in a 6DoF model using a positioning algorithm. The inverse Dynamics Virtual Domain (IDVD) technique was used to generate real-time trajectories for dynamic soaring applications using a point mass model of a powered unmanned sailplane and the fuel saving trajectories were validated using a high fidelity 6DoF model and a classical controller. An important outcome of the research is the fact that energy saved during dynamic soaring flight was also realized due to a sinusoidal manoeuvre using reduced thrust. In this manoeuvre the kinetic energy is converted into potential energy by gaining altitude and by reducing airspeed. Then initial values of altitude and speed are gained by loosing the altitude. In this process a horizontal distance is travelled by using reduced thrust.