Collision Avoidance for Fixed-Wing Unmanned Aerial Systems Using Morphing Potential Field Navigation with Robust and Predictive Control

Abstract

The vast utility of unmanned aerial systems in wide-ranging applications, whether civil, militaristic, or academic, has accelerated the prospect of integration into civil airspace, and further, use in proximal collaborative operations and congested urban areas. For safety in autonomous operations of the kind, onboard navigation algorithms will require the ability to generate collision free paths in real-time. Guidance and robust control systems must maintain tight spatial and temporal tracking of these paths in the presence of environmental hazards, a twofold objective involving both sensors and control system development; here, the latter is targeted. This work proposes a combined real-time navigation, guidance, and robust control scheme for collision and obstacle avoidance in the particular case of fixed-wing unmanned aerial systems operating in demanding proximal and congested settings. Classical artificial potential field navigational approaches are uniquely reformulated, morphing the fields by consideration of six-degrees-of-freedom dynamic characteristics and constraints of high speed and high inertia fixed-wing aircraft. Time-varying waypoints are planned in a predictive horizon and subsequently embedded into an integrated guidance and nonlinear model predictive controller. Nonlinear six-degrees-of-freedom simulation of a suite of vehicle-to-vehicle and obstacle avoidance scenarios in unstructured environments demonstrate robust, real-time, and efficient avoidance capabilities of the developed algorithms.