基于k-shell分解的多智能体牵制控制算法  

Multi-agent pinning control algorithm based on k-shell decomposition

作　　者：何明 许元云 刘锦涛 周波 丁晓晖 HE Ming;XU Yuan-yun;LIU Jin-tao;ZHOU Bo;DING Xiao-hui(Command&Control Engineering College,Army Engineering University of PLA,Nanjing 210007,China;Unit 94860 of PLA,Nanjing 210000,China)

机构地区：[1]解放军陆军工程大学指挥控制工程学院,南京210007 [2]解放军94860部队,南京210000

出　　处：《控制与决策》2020年第10期2556-2560,共5页Control and Decision

基　　金：国家重点研发计划项目(2018YFC0806900);中国博士后科学基金项目(2018M633757);江苏省重点研发计划项目(BE2016904,BE2017616,BE2018754,BE2019762);江苏省博士后科学基金项目(2019K185)。

摘　　要：针对多智能体网络在牵制控制过程中存在的网络分裂现象,考虑到牵制节点选择对多智能体收敛速度的影响,提出一种基于k-shell分解的牵制控制算法.首先根据节点连通度划分子网;然后提出基于k-shell分解的牵制节点选择方法;最后完成多智能体的牵制控制.理论推导证明,采用该算法后整个智能体网络最终将形成一个子网.分析对比3种牵制控制算法,通过实验仿真结果验证所提出算法能够实现多智能体的一致性,有利于提高多智能体的收敛速度.Aiming at the network splitting phenomenon in the control process of multi-agent network, considering the influence of informed agent selection on the convergence speed of multi-agents, the pinning control algorithm based on k-shell decomposition is proposed. Firstly, the subnet is divided according to the node connectivity. Then, the method of selecting the informed agent based on k-shell decomposition is proposed. Finally, the pinning control of multi-agent is completed. Theoretical derivation proves that after the algorithm is adopted, the entire mutli-agent network eventually form a connected graph. The experimental results verify that the proposed algorithm can achieve the consensus of multi-agent, and benefits to improve the convergence speed compared with three pinning control algorithms.

关 键 词：多智能体 牵制控制 一致性 k-shell分解 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]