Probabilistic Roadmaps And Ant Colony Optimization For UAV Mission Planning

This paper presents a two-tiered approach to mission planning based on Probabilistic Roadmaps for path planning and Ant Colony Optimization for task planning. The path planner searches paths in a finite, obstacle-constrained three dimensional space whereas the task planner determines a near optimal task order for a set of tasks and a group of heterogeneous agents. The mission planning system is employed in an unmanned helicopter project that researches on intelligent technologies for small, low-altitude aerial vehicles. Results from simulation experiments demonstrate the feasibility of combining a dedicated path planner with a task planner