摘要
针对复杂非结构环境下作业任务对机器人的多功能性、柔性化等特殊需求,提出一种以映射矩为目标构型识别依据、矩阵元素为成型运动核心的晶格式集群机器人矩阵成型方法.基于离散化处理后的晶格式群体系统与元素离散排列、对称分布的矩阵的相似性,在预处理阶段,将用户给定的二维目标几何构型任务进行矩阵二值化处理后,与初始构型矩阵嵌套和匹配,并依次通过矩阵映射运算、初始构型体的相对定位实现集群机器人对目标几何构型的有效识别.在成型阶段,晶格式机器人个体以矩阵元素为运动依据、晶格单元为成型载体,设计晶格式集群机器人的个体自主成型规则,有效解决群体系统中"走哪里"和"怎么走"的问题.最后,通过仿真实验表明群体矩阵成型方法的可行性和有效性.
Aiming at the special requirements in non-structural environment, a matrices shape formation approach for swarm robots is presented, which includes goal shape recognized based on mapping matrices and formation movement based on matrix elements, to adapt the versatility and flexibility of robots in complex tasks. Based on the similarity between the discretized crystal format group system and the matrix with discrete arrangement and symmetrical distribution of elements, during the process of pretreatment, the 2-D user-specific shape is binarized as a goal matrix, which is matched with the matrix of the initial shape. Through matrix mapping operations and the relative localization approach of the initial shape, the goal shape is recognized by swarm robots. During the process of formation movement, a movement rule to solve the problem of "where to go" and "how to go" is proposed according to the movement feature based on the matrix element and the formation carrier based on the lattice unit. Finally, the feasibility and validity of the matrix shape formation approach are verified by simulation experiments, and following the shape formation algorithm, after a novel modular robot is designed and a latticed formation platform for swarm robots is built, the nine latticed robots by two kinds of experiments are completed.
作者
杨宏安
段鑫
张昭琪
曹帅
昝文佩
YANG Hong-an;DUAN Xin;ZHANG Zhao-qi;CAO Shuai;ZAN Wen-pei(Electromechanical College,Northwestern Polytechnical University,Xi’an 710072,China)
出处
《控制与决策》
EI
CSCD
北大核心
2020年第10期2391-2398,共8页
Control and Decision
基金
国家自然科学基金项目(51775435)
西北工业大学种子基金项目(ZZ2019091)。
关键词
集群机器人
矩阵成型
晶格式移动机器人
个体成型规则
实验验证
swarm robots
matrix shape formation
latticed robot
individual movement rules
experiment verification