Multi-robot System Self-repairing Algorithm Based On 2-hop Neighbor Information Diffusion

In recent years,the working environment of robots and the tasks that need to be completed have become increasingly complex.Multi-robot systems have good application prospects because of their characteristics of better scalability,redundancy,economy,and robustness.It is unavoidable that environmental disturbances cause multiple robots in the system to fail at the same time since the working environment of multiple robots is a complex industrial environment or a wild environment.And this could directly change the topology of the multi-robot network and cause the network connectivity and motion synchronization of the system to decrease,which result in the phenomenon that the efficiency of system tasks is affected,and even the tasks can not be completed.Therefore,multi-robot systems require an algorithm that can recover network connectivity and motion synchronism when multiple robots failed at the same time.The distributed selfhealing algorithms proposed by the predecessors for the synchronization index cannot repair the multi-robot network that is not connected due to the lack of robots,and they cannot repair multiple failed robots at the same time.Aiming at the problem that the information between non-neighbor robots can not be shared in distributed control,the disconnected multi-robot network can not be repaired,and the multi-robot missing can not be repaired,this paper proposes a self-healing algorithm based on 2-hop neighbor information diffusion.The robot only needs to perform local information interaction with neighbors to complete the task of repairing each failure robot.At the same time,aiming at improving the connectivity and synchronization of multi-robot network globally,this paper also introduces gradient information,and designs a new gradient generation,diffusion,and update mechanism to ensure that distributed handover topology control can achieve global connectivity and improved synchronization in multi-robot networks.Furthermore,the stability of the gradient generation and diffusion algorithm is analyzed in this paper.It is proved that the proposed robot selection rule can improve the multi-robot network connectivity and motion synchronization.And the simulation and experiment results validate the convergence and stability of the self-healing algorithm.In order to verify that the distributed self-healing algorithm proposed in this paper can improve the performance when there are multiple robot failures,this paper specially builds a simulation and experimental platform for multi-robot system and launched a number of tests.Firstly,in the simulation platform,this paper tests the self-repair results of robots with different robot system scales or different network topology.Further,this paper tests the self-healing results of multiple robots failure under different network topology on the experimental platform.The simulation and experimental results verify that the proposed algorithm not only repairs the multi-robot network topology after multirobot failed,but also improves the connectivity and motion synchronization of multi-robot networks.