Abstract

The main focus of this thesis is to provide strategies for collision-free motion in multi-swarm coverage control. Motivated by the diverse use of agent-based systems for various tasks, the scenario of multiple non-cooperating swarms independently covering a common area is presented. Using Voronoi tessellation in coverage control, collision-free motion of agents within the same swarm has been proven before. However, in the case of multiple swarms following their own objectives, these guarantees do not hold. To address this issue, the Optimal Reciprocal Collision Avoidance (ORCA) method used for safe navigation in multi-agent scenarios is applied to multi-swarm coverage control. Assuming knowledge regarding the positions of all agents, the proposed methodology is formally analyzed for planar motion and validated through Monte Carlo simulations. Subsequently, the collision-avoidance algorithm is investigated in environments where bounded disturbance measurement uncertainties are present. To account for these disturbances, an extension of ORCA is proposed. Formal guarantees are presented for motion without collisions between agents. This is done under the assumption that the input needed to counteract the disturbance can always be achieved. The theoretical results are applied to coverage control of multiple non-cooperating swarms and validated through MATLAB simulations.