引用本文:付瑛博,杨紫雯,朱善迎,陈彩莲.利用方位角观测的多机器人系统离散分布式最优围陷控制[J].控制理论与应用,2022,39(10):1815~1824.[点击复制]
FU Ying-bo,Yang Zi-wen,ZHU Shan-ying,Chen Cai-lian.Discrete-time distributed optimal entrapment for multi-robot systems using bearing measurements[J].Control Theory and Technology,2022,39(10):1815~1824.[点击复制]
利用方位角观测的多机器人系统离散分布式最优围陷控制
Discrete-time distributed optimal entrapment for multi-robot systems using bearing measurements
摘要点击 2026  全文点击 474  投稿时间：2021-10-04  修订日期：2022-09-30
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/CTA.2022.10943
  2022,39(10):1815-1824
中文关键词  多机器人系统  离散时间  纯方位测量  优化增益
英文关键词  Multi-robot system  Discrete-time  Bearing-only measurements  Optimal gains
基金项目  国家自然科学基金（61922058，62103277，62025305）和中国博士后研究基金创新人才计划（BX2021181）
作者单位邮编
付瑛博 上海交通大学 200240
杨紫雯 上海交通大学 
朱善迎* 上海交通大学 200240
陈彩莲 上海交通大学 
中文摘要
      本文研究了离散时间多机器人系统对静态目标的同步定位与围陷问题。首先，为了引导多机器人系统以任意编队在任意轨道上运动，本文提出了基于虚拟系统法的路径规划方案。其次，为了实现在缺少机器人全局位置信息下的目标定位与跟踪问题，本文设计了基于方位角观测的离散时间估计器与控制器的联合设计方法。估计 器充分利用了正交性质估计实时相对位移，并基于估计值设计控制律。基于离散时域拉塞尔原理与迭代法，本文给出了估计误差与跟踪误差收敛的充分条件与稳定性证明。此外，本文还设计了一种增益调整优化算法以提高收敛速率，使得离散时间多机器人系统可以更快达到理想编队。最后，仿真实验结果证明了整体方案的有效性。
英文摘要
      In this paper, we investigate the problem of simultaneous localization and entrapment of a static target by robots with discrete-time dynamics. Firstly, to lead the multi-robot system to move on an arbitrarily shaped orbit in an arbitrarily shaped formation, a scheme for path planning based on the virtual system is presented. Secondly, in order to locate and entrap the target without prior information of the robots’ global position, a joint discrete-time estimator-controller using relative bearing-only measurements is proposed. The estimator is designed by exploiting the orthogonality property and based on this estimator, the controller is developed. Using LaSalle’s theorem for the discrete systems and iterative method, we propose the sufficient conditions under which the convergence of estimation error and tracking error is proved. Moreover, to improve the overall performance, an optimal gains tuning method is developed, so that discrete-time robots can achieve the desired formation at a faster convergence speed. Finally, simulation results are presented to verify the effectiveness of the overall scheme.