Autonomous hierarchical control of multiple unmanned combat air vehicles (UCAVs)

In this paper we present a hierarchical control scheme that enables multiple UCAVs to achieve demanding missions in hostile environments autonomously. The objective is to use a swarm of UCAVs for a SEAD type mission: fly the UCAVs in a formation to an enemy territory populated with different kinds of threats, collect enemy information or destroy certain targets, and return to the base, all without human intervention. The scheme is an integration of four distinct components, including: (1) high level Voronoi diagram based path planner to avoid static threats; (2) low level path planner to avoid popup threats; (3) differential flatness based trajectory generator to generate dynamically feasible trajectory; and (4) semi-globally stable formation control algorithm to maintain the formation. The scheme was implemented in Matlab and demonstrated very effective path planning, trajectory generation, and formation flying capabilities. We also developed an interface from Matlab to IWARS, a high fidelity battlefield simulation environment developed by Boeing. This enabled us to study the effectiveness of our scheme under various battle scenarios using IWARS.