Simulation analysis of autonomous aerial refueling procedures

An approach for simulating an autonomous aerial refueling maneuver based on optical sensors is presented. The primary purpose of the simulation model and the simplified control system design is to derive requirements for different components contributing to the system. A nonlinear simulation model of an unmanned, slightly unstable receiver aircraft, including sensor and actuation dynamics, is used as basis for the study. The refueling basket is modeled as a simplified dynamic system subject to atmospheric disturbances causing basket motion superposed by stochastic deformations. A flight control system based on linear methods is designed to stabilize the receiver aircraft, with bank angle and angle of attack as command variables. Path control is performed by nonlinear dynamic inversion of a simplified point mass model of the receiver aircraft. Trajectory control of the probe nose is implemented based on linear control methods, utilizing position estimates for the probe relative to the basket provided by Kalman filtering and successive geometric transformation. For the input sensor signal for the estimates, the location of marker points attached to the basket in the focal plane of a receiver aircraft mounted camera is used. The performance of the approach is assessed by nonlinear simulation, focusing on the identification of sensitive parameters influencing the refueling procedure.