An active controller for a UAV helicopter carrying a slung load is described in this paper. The objective of the controller is to allow the UAV to safely transport a slung load and to place it precisely on a moving ground platform such as a moving truck or a ship. In order to fulfill this objective, the active controller is synthesized in terms of three sub-components; first a target position tracker which generates position tracking commands, second a load oscillation controller which generates load oscillation damping commands, and third an adaptive neural network which compensates for uncertainties associated with flight environment and/or modeling errors. A linear PD controller is used for the target position tracking control. A nonlinear controller based on feedback linearization of the slung load dynamics is used for the load oscillation control. A single hidden layer neural network with an adaptive gain update is used for uncertainty compensation. The proposed controller is evaluated in simulations within the Georgia Tech UAV Simulation Tool (GUST) and in flight tests using the GTMax UAV helicopter testbed. Both simulation and flight test results are presented to demonstrate the effectiveness of the proposed controller in dampening of load oscillations while simultaneously reducing position errors relative to a virtual moving ground platform, in the presence of random ground vehicle motion, wind gusts and modeling errors.
|Original language||English (US)|
|Number of pages||10|
|Journal||Annual Forum Proceedings - AHS International|
|State||Published - Dec 1 2010|
|Event||66th Forum of the American Helicopter Society: "Rising to New Heights in Vertical Lift Technology", AHS Forum 66 - Phoenix, AZ, United States|
Duration: May 11 2010 → May 13 2010
All Science Journal Classification (ASJC) codes