Distributed Rendezvous of Heterogeneous Robots with Motion-Based Power Level Estimation

Abstract

This article introduces distributed rendezvous algorithms to make all heterogeneous robots (agents) rendezvous at a designated agent in environments with many obstacles. The proposed control laws are developed considering heterogeneous agents having distinct power level, sensing range, or communication range. It is assumed that each agent is initially distributed in a cluttered environment. An agent cannot utilize global positioning systems (GPS) to localize itself. Each agent can only sense neighboring agents and moves based on proximity interaction. This article presents distributed rendezvous algorithms so that heterogeneous agents rendezvous at a designated agent while maintaining (directed) network connectivity. The speed of each agent is set proportional to the remaining power level of the agent. This article then presents a method of estimating the power of each agent by measuring the movement of the agent. Then, we present a method of sharing power if an agent is lack of power. Under the proposed approach, we achieve distributed rendezvous as well as power recharging. It is proved that multiple heterogeneous agents rendezvous at the designated agent while maintaining (directed) network connectivity. Utilizing MATLAB simulations, the performance of the proposed distributed controls is verified in an environment with many obstacles.