Energy-aware Persistent Control of Heterogeneous Robotic Systems

In this paper, we consider the problem of online planning and control of heterogeneous robotic systems for energy constrained safety operations. Specifically, a team of aerial and ground vehicles is tasked with completing a set of previously unknown goals within a bounded region and automatically deciding future recharging operations to avoid depleting energy throughout a mission. We propose a prioritized distributed virtual potential scheme with spring-mass and magnetic virtual force control behaviors in order to plan and coordinate the motion of the vehicles so that to guarantee positive energy (achieved with recharging operations) and task completion. The proposed approach is demonstrated to work also in the event of failures and eventually self-heal the heterogeneous system over time. Stability analysis in the sense of Lyapunov is presented to prove safety of the system. Finally, we validate the proposed approach with simulations and experimental results.