The Human-robot Coordination Algorithms Of Teleoperated Multi-robot Exploration Systems Under Connectivity Constraints

Due to the advances of the communication technology, multi-robot systems havebeen considerably developed. Environment exploration is an important issue in thefeld of multi-robotics, and have broad applications in the felds of rescue and mili-tary. Human has the strategic vision and the analysis ability that multi-robot systemsdon’t have, but most of the existing environment exploration approaches enable themulti-robot system to explore autonomously and do not enable human to intervenein the exploration so as to exert human’s advantages and improve exploration perfor-mance. Sothekeyofmulti-robotexplorationistocoordinateintentionsoftheteleoper-ator with the task allocation decisions of transmission-bandwidth-and-energy-limitedmulti-robot systems under the premise of connectivity constraints.This paper focuses on two typical application scenarios: Wide-Range-RoamingExploration and Small-Range-Specifc Exploration, collaborative exploration betweena teleoperator and a multi-robot system is discussed and Human-Robot CoordinationAlgorithms of Teleoperated Multi-Robot Exploration Systems under a star network aswellasanAd-hocnetworkaredesignedandimplemented. Thetransmitbandwidthandpower consumption is reduced and enable the human to intervene in the explorationunder the premise of satisfying the connectivity constraints. The contributions are asfollows:1. The system model of the Teleoperated Multi-Robot Exploration System under astar network is composed of a Mobile Base Station and several Robot Explorers.First, a Teleoperator Interest Oriented Task Allocation algorithm under a star network is designed in order to enable the multi-robot system to allocate thetasks according to the operator’s interest efectively and harmonically. Further,a robot Movement Trade-of Method is proposed: The optimization index of themethod balances the Obedience, Connectivity and Explorability of the system.The problem is solved to generate the fnal movement of the mobile base stationand the robot explorers.2. Thesystem modelof theTeleoperatedMulti-RobotExploration SystemunderanAd-hoc network is composed of several robot explorers. By adapting the Tele-operator Interest Oriented Task Allocation algorithm under a star network into amethod that could be used under an Ad-hoc network, the multi-robot system isable to coordinate its own task allocation decisions with the operator’s interest.Further, by considering the importance of each multi-robot’s task and solving aMaximized Vertice Weight Sum Steiner-Tree Problem, the topology structure isrestored, while the explorations to the operator’s regions of interest are ensured.Finally,byusingabreadth-frstsearch,adatafusioncenterisdynamicallychosento ensure the efciency and the robustness of the data transfer.3. Finally,theHuman-RobotCoordinationAlgorithmsofTeleoperatedMulti-RobotExploration Systems under a star network as well as an Ad-hoc network areimplemented and tested by both simulations and experiments, including com-parisons of the average exploration time cost, standard deviation of the explo-ration time cost, efective exploration efciency ratio of the designed systemwith those of the fully autonomous methods. The transferred data amount andthe exchangeddataamountintheon-site multi-robotnetworkarealsocompared.Results show that the designed algorithms enable the Teleoperated Multi-robotExploration Systems have higher exploration efciency, data transfer efciencyand lower transmit power consumption.