Using stream functions for complex behavior and path generation

Abstract

A methodology for generating complex trajectories for vehicles in an uncertain, adversarial, dynamic environment is presented. Stream functions are used as a basis for generation becuase they simply and quickly produce trajectories around obstacles. Streamline theory is presented, followed by examples of complex trajectory generation, including avoidance of multiple obstacles, planning with regard to risk, formations of vehicles, and various tactical maneuvers. Examples are demonstrated in a high-fidelity battlefield simulator with tanks, airplanes, surface-to-air missile sites, and unmanned air vehicles as the objects of control. © 2003 by Jeffrey M. Sullivan.