Abstract
The major technological obstacle to be overcome for practical and reliable autonomous probe-and-drogue aerial refueling is obtaining accurate relative position and attitude measurements during the docking phase. An integrated controller-sensor-navigation system for this task must be robust and possess good disturbance rejection properties. Previous attempts to solve this problem have used video servoing with pattern recognition algorithms and the differential Global Positioning System. This thesis seeks to determine the feasibility of autonomous aerial refueling by developing a robust docking controller and integrating it with the relative position and attitude measurements from a novel Vision-based Navigation (VisNav) sensor. VisNav accurately determines the line of sight vector between a positioning sensing diode and a target configured with multiple light emitting diode beacons. A study is conducted to determine the best number and placement of the beacons on the drogue and the best location to mount the sensor on an Unmanned Air Vehicle (UAV). Optimal Nonzero Set Point and optimal Command Generator Tracker controllers are developed and used to simulate six degree-of-freedom docking maneuvers using dynamical system models of a UAV and a refueling drogue. Test cases for stationary and moving drogues in atmospheric turbulence are evaluated in terms of docking position errors, control effort, control rate, and quadratic cost. Simulation results demonstrate that a Proportional Integral Filter Command Generator Tracker controller, coupled with the VisNav sensor and navigation system, provides a viable candidate solution to the autonomous aerial refueling problem. The beacon lights can be placed in the location of lights currently on the drogue, and the sensor can be placed at the base of the refueling probe on the UAV.
Kimmett, Jennifer Jones (2002). Autonomous aerial refueling of UAVS utilizing a vision based navigation system. Master's thesis, Texas A&M University. Available electronically from
https : / /hdl .handle .net /1969 .1 /ETD -TAMU -2002 -THESIS -K48.