基于合作自主定位的群体模式自动生成方法  

Automatic Swarm Pattern Generation Method Based on Cooperative Autonomous Localization

作　　者：范衠 王诏君 石泽 王琛 王柳 蔡堉伟 黄华兴 宁为博 董朝晖 李文姬 朱贵杰[1,2] 李恪 李兵[1,2] 谷敏强 朱晓敏 包卫东[4] FAN Zhun;WANG Zhaojun;SHI Ze;WANG Chen;WANG Liu;CAI Yuwei;HUANG Huaxing;NING Weibo;DONG Zhaohui;LI Wenji;ZHU Guijie;LI Ke;LI Bing;GU Minqiang;ZHU Xiaomin;BAO Weidong(Department of Electronic Engineering,Shantou University,Shantou 515063,Guangdong,China;Key Lab of Digital Signal and Image Processing of Guangdong Province,Shantou University,Shantou 515063,Guangdong,China;Department of mathematics,Shantou University,Shantou 515063,Guangdong,China;College of Systems Engineering,National University of Defense Technology,Changsha 410000,Hunan,China)

机构地区：[1]汕头大学电子系,广东汕头515063 [2]汕头大学广东省数字信号与图像处理技术重点实验室,广东汕头515063 [3]汕头大学数学系,广东汕头515063 [4]国防科技大学系统工程学院,长沙410073

出　　处：《汕头大学学报（自然科学版）》2021年第2期14-28,共15页Journal of Shantou University：Natural Science Edition

基　　金：中央军委科技委基础研究项目(18-163-11-ZT-003-008-02);中央军委科技委基础研究项目(193-A14-226-01-01)。

摘　　要：传统的基因调控网络(Gene Regulatory Network,GRN)模型假设群体机器人具有对环境的全局认识能力,这就要求群体机器人所携带的传感器具有很强的性能.在实际中,由于群体机器人的传感器误差和外部干扰,群体机器人很难获得所在环境的全局信息.为此,我们提出一种基于合作自主定位的群体模式自动生成方法.该方法的优点是每个机器人事先在不知道自身和邻居的位置前提下,通过传感器测量出自身和邻居之间的距离,再利用三点定位方法确定群体机器人位置.所提出的方法是由一个三层基因调控网络模型所构成,即第一层负责群体机器人协作自组织定位.第二层是针对给定环境的自适应生成关于目标的群体模式.第三层是一个无中心化、分步式的控制机制,它驱动群体机器人机动到由第二层产生的群体模式上.最后,通过一些实验验证所提出的方法的有效性.The traditional Gene Regulatory Network(GRN)model usually assumes that the swarm robot has global awareness of the environment,which requires strong performance of the sensors carried by the swarm robot.In practice,due to the errors of sensors and the external disturbance of the swarm robot,it is difficult for the swarm robot to obtain accurate global information.Therefore,an automatic swarm pattern generation method based on cooperative autonomous localization is proposed.The advantage of this method is that each robot can measure the distance between itself and its neighbors by sensors in advance without knowing the positions of itself and its neighbors,and then uses a three-point localization method to determine the swarm robot positions.The proposed method has a three-layer gene regulation network model,i.e.,the first layer is responsible for the collaborative self-organized localization of the swarm robots.The second layer is to generate adaptive patterns for the swarm robot in given environments.The third layer is a decentralized,stepwise control mechanism that drives the swarm robot to maneuver onto the swarm patterns generated by the second layer.Finally,the effectiveness of the proposed approach is verified by a set of experiments.

关 键 词：群体智能机器人 基因调控网络模型 目标围捕 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]