Potential flow field navigation with virtual force field for UAS collision avoidance

摘要

This paper presents the combination of Potential Flow Field (PFF) and Virtual Force Field (VFF) methods to construct a viable means for UAS's to autonomously navigate and avoid collisions with obstacles. A validated flight dynamics model of a fixed-wing UAS was integrated with a heading hold autopilot that receives commanded heading angles from the selected output of the PFF and VFF methods. PFF primarily provides real-time heading angle commands derived from a velocity vector field. VFF provides collision avoidance heading angle commands when it calculates and determines that the UAS is in danger of collision with an obstacle. The combination of the two methods resulted in successful collision-free navigation in simulation through multiple scenarios including single obstacle, multiple obstacles, maze, and urban settings.