基于梯度扩散机制的多机器人网络故障修复算法研究

Repair algorithm for multi-robot network failures with gradient diffusion mechanism

在多机器人网络中,机器人节点发生故障会导致原有的通信网络遭到破坏,文中提出了一种基于梯度扩散机制的故障修复算法,使机器人能够通过局部的信息交互递归地修复网络拓扑。首先通过梯度生成和扩散机制在多机器人通信系统中产生稳定的梯度分布,然后根据选举规则选出最佳修复路径。此外,机器人节点额外维护了二跳邻居节点表和最佳邻居机器人信息,用来解决多个机器人同时发生故障引起的网络中断、修复路径冲突等问题。经过仿真验证,该算法能够有效地恢复多机器人故障系统的网络拓扑,保证多机器人网络的连通性和同步性,同时减少了故障网络的修复时间。

In the multi-robot network,the failure of the robot node will destroy the original communication network.This paper proposes a fault repair algorithm based on the gradient diffusion mechanism.The algorithm enables the robot to recursively repair the network topology through local information interaction.Firstly,a gradient distribution is generated in the multi-robot communication system by the gradient generation and diffusion mechanism.Secondly,the optimal repair path is elected according to the election rules.In addition,the robot node additionally maintains 2-hop neighbor node table and optimal neighbor robot information,to solve network interruption and repair path conflict caused by simultaneous failure of multiple robots.The simulation proves that the algorithm can effectively restore the network topology of the multi-robot fault system,ensure the connectivity and synchronization of the multi-robot network and reduce the repair time of the faulty network.