Safe Voronoi-Based Coverage Control for Multi-Robot Systems with Constraints

• André Chaves Magalhães UFMG
• Guilherme Vianna Raffo UFMG
• Luciano Cunha A. Pimenta UFMG

Resumo

This paper presents a control strategy based on Voronoi partitions to cover an environment with a multi-robot system with state and control constraints. A bounded convex work region is partitioned at each instant using a Voronoi algorithm, providing agents with non-overlapping navigation zones. The algorithm incorporates individual weights for each agent, modifying the partitions to assign a larger area to the agent with the highest weight. The proposed deployment aims to bring the entire system to a stable static configuration corresponding to optimal region coverage. This objective is achieved using a model predictive control and artificial vector fields to guide agents to curves with center points placed at Chebyshev Configurations of the Voronoi partition. The agents are modeled as Dubins airplanes with constant speed and curvature restriction. Simulations are performed using MATLAB software, and the results show the efficiency of the proposed control strategy.

Palavras-chave: Model Predictive Control, Dubins Vehicles, Coverage Control, Chebyshev Center, Vector Fields