🤖 AI Summary
To address coverage gaps and communication disconnections caused by robot offline states during charging in energy-constrained multi-robot systems, this paper establishes a hybrid system model integrating motion dynamics, energy evolution, and network topology. We propose an energy-aware bearing-rigidity-based network design framework. A tri-modal coordination mechanism—encompassing coverage, return-to-base, and charging phases—coupled with energy-constrained guardian conditions ensures structural rigidity and topological connectivity throughout the charging process. Crucially, we embed real-time energy states into the evolution of bearing-rigid formations, enabling dynamic connectivity maintenance and persistent high-quality coverage. Numerical simulations demonstrate that, while strictly respecting individual energy constraints, the method significantly improves coverage completeness (≥23% increase) and network connectivity persistence (89% reduction in disconnection duration).
📝 Abstract
The problem of multi-robot coverage control becomes significantly challenging when multiple robots leave the mission space simultaneously to charge their batteries, disrupting the underlying network topology for communication and sensing. To address this, we propose a resilient network design and control approach that allows robots to achieve the desired coverage performance while satisfying energy constraints and maintaining network connectivity throughout the mission. We model the combined motion, energy, and network dynamics of the multirobot systems (MRS) as a hybrid system with three modes, i.e., coverage, return-to-base, and recharge, respectively. We show that ensuring the energy constraints can be transformed into designing appropriate guard conditions for mode transition between each of the three modes. Additionally, we present a systematic procedure to design, maintain, and reconfigure the underlying network topology using an energy-aware bearing rigid network design, enhancing the structural resilience of the MRS even when a subset of robots departs to charge their batteries. Finally, we validate our proposed method using numerical simulations.