Swarming Coordination with Robust Control Lyapunov Function Approach

A novel approach to address swarm coordination of a multi-agent system consisting of Unmanned Aerial Vehicles (UAVs) using Robust Control Lyapunov Function (RCLF) is proposed in this paper. The approach addresses swarm coordination by commanding each agent to reach a desired separation distance with its neighbors and achieving heading consensus. The swarming is modeled as a decentralized scheme using behavioral approach and the inter-agent interaction is described with sigmoid artificial potential field. For this engineering swarm to mimic the natural flocking of birds, the local interaction is also modeled with Voronoi partitions method such that each agent only interacts with their immediate neighbors. Mid-air collision can be avoided by specifying a suitable desired separation distance. The simulation results show that the approach yields less intense artificial potential field, and is able to achieve the three main criteria of Reynolds' boids model (cohesion, separation, and alignment) effectively. Agents are able to satisfy the behavioral criteria in following an unknown informed agent (leader) in the presence of disturbances. Experimental results using quadrotor test-beds are presented to demonstrate the feasibility of the decentralized behavioral scheme.