Description

Nonlinear Controller on a Fixed Wing Aircraft by Frant Sobolic and Saif Al-Hiddabi

Unmanned aerial vehicles (UAVs) are increasingly being used for practical purposes such as reconnaissance, surveillance, and search and rescue missions because of their ability to perform aggressive maneuvers in highly constrained areas while maintaining stability and control. Most UAVs fall into two main types of classes: vertical take-off and landing (VTOL) and fixed wing aircraft. Vehicles, such as helicopters and quadrotors are often used in highly constrained missions but are hindered by their slow transition from hover to translational flight. Another type of UAV, such as the Predator, has a fixed wing configuration that allows the aircraft to maneuver quickly in a translational manner but is usually constrained to fly at speeds above stall. By combining the low speed maneuverability of a VTOL vehicle with the quick translational abilities of a fixed wing UAV, a single vehicle can be designed to perform well in all flight regimes. Such a vehicle would be capable of performing both a perch style landing and a traditional landing. Previous vehicles that have demonstrated this type of capability combine the actions of two independent controllers that perform during either hover or translational flight. The problem statement then becomes a matter of intellectual transitioning between the hover and translational flight regions of the two controllers but through the use of a nonlinear controller, these complexities can be embraced in a single design.

This research presents a nonlinear control design that is capable of performing in both of these flight regimes. It is a nonlinear Lyapunov back stepping controller that follows smooth time varying position commands, as well as constant position commands. The controller design has been established and the implementation procedure has begun. The vehicle choosen to serve as the testbed for this control design is the high performance Clik indoor aerobatic plane shown below. The vehicle sensing (position, velocity and angular rate) are all done off-board in the Real-time indoor Autonomous Vehicle test Environment (RAVEN) which provides all the information needed to fuel this controller.