Dynamic soaring is a technique by which wind gradients are utilized to extend flight times. This was originally observed amongst birds like albatrosses and has been identified to have huge potential in the flight of high-endurance UAVs. Earlier works have focussed on generating dynamic soaring trajectories including periodic orbits through various methods like Gauss-Pseudo spectral method, interior-point method,
and numerical integration methods. Stability of dynamic soaring orbits is important since the trajectories can get disturbed by a gust or crosswinds causing the UAV to veer off-course. Although a control system can be designed, an open-loop stable orbit can reduce the control effort and power requirement. In this paper, the problem of
studying stability is treated from the context of a periodic coefficient system. Variation of Phugoid damping values with wind shear has been studied. For assessing the dependence of the stability of dynamic soaring orbits on wind shear (or friction velocity) a Monte-Carlo based approach is used. In addition a stability augmentation system, based on earlier work on control of linear periodic coefficient system, has
been proposed. A possible application of Dynamic Soaring, for navigating through designated waypoints (targets on the ground) has also been proposed. Such an application finds benefit in surveillance applications. Here we attempt dynamic soaring trajectory generation for passing through waypoints. This method can be useful for
surveillance of designated points. Additionally, unsteady effects on dynamic soaring trajectories has been explored. A UVLM
code for getting the forces acting on a full aircraft has been developed in MATLAB, in what is believed to be the first of its kind. The code has been thoroughly tested using standardised tests like TR-1208 and verified. It was seen that the unsteady forces deviate to a large extent from the constant coefficient system’s forces and a numerical integration of the trajectory in MATLAB causes divergence that couldn’t be stabilised even by a Linear Quadratic Regulator (LQR) controller.
Trajectories have been generated for soaring of UAV’s along hill slopes with suitable perturbation in wind velocity given in the x-direction. The wind shear for the perturbed wind profile was calculated using concepts of modified Bessel functions. The resulting trajectory was also stabilised using a Pole-Placement and LQR approach. Soaring of UAVs along ridges, at high speeds, has great sporting interest. So far, such an event is achieved by intuitive and trial and error means. Here trajectory generation for soaring at high speeds for a 6 DOF UAV model has been demonstrated and its stabilization using LQR has been shown. The control inputs thus achieved can be tested on a real UAV for
achieving high-speed soaring. This can be seeds of future work where a more detailed approach can be arrived at for making feasible soaring at high speeds.