Development of a novel method for autonomous navigation and landing of unmanned aerial vehicles

In this work, control techniques for the autonomous navigation and landing of an Unmanned Aerial Vehicle (UAV) are developed and compared. Controllers were developed and implemented on two different aircraft models: the Lockheed-Martin F-16 and AAI Corporation/Israel Aircraft Industries RQ-2 Pioneer. Due to the expense of modifying the pre-existing F-16 flight control system, the controller is implemented outside of the closed loop. Proportional-integral-derivative and proportional-integral controllers are developed for holding the aircraft at a desired velocity and altitude. The aircraft are approximated as Dubins vehicles constrained to travel on a two-dimensional surface for decreased simulation time. Using the simplified model two control techniques are developed and then compared. The first uses a proportional feedback controller based on the Rhumb-line that the aircraft is traveling along. The second control technique uses a trajectory determined from an algorithm using the Dubins path determination for the shortest travel distance between two points. A sliding mode controller is developed to guide the simplified model along the Dubins path trajectory. The advantage of the Dubins path trajectory is that it allows for a closed-form time estimate to reach the desired way-point. Comparison between the two navigation techniques using the simplified system shows a significant decrease in time to way-point for the Dubins curve trajectory controller. The Rhumb-line controller and a hybrid Rhumb-line/Dubins path controller are implemented on nonlinear models of both aircraft. Simulation of both controllers on the nonlinear model shows acceptable performance in guiding the aircraft between way-points. Also, the time to way-point for the nonlinear aircraft model guided by the hybrid controller is within 5% of the closed-form Dubins trajectory estimate. Autonomous landing is accomplished utilizing the path guidance and altitude controllers. The nonlinear simulated aircraft successfully followed the glideslope from way-point to runway.