自重构模块化机器人路径规划方法综述  

Path planning for self-reconfigurable modular robots: a survey

作　　者：刘永进[1,2] 余旻婧 叶子鹏 王昌凌 

机构地区：[1]清华大学计算机科学与技术系,中国北京100084 [2]清华信息科学与技术国家实验室,中国北京100084 [3]Department of Design Engineering&TU Delft Robotics Institute,Delft University of Technology,Delft 2600 AA,The Netherlands

出　　处：《中国科学：信息科学》2018年第2期143-176,共34页Scientia Sinica(Informationis)

基　　金：国家重点研发计划(批准号：2016YFB1001200); 国家自然科学基金创新研究群体项目(批准号：61521002)资助

摘　　要：自重构模块化机器人指由一系列通用模块组装而成、能够根据所处环境或任务的变化依靠模块间的相互通信和自主运动重组为另一种适应新环境、新任务构型的机器人,对它的研究近年来已成为学术界和工业界关注的热点.模块化机器人的自重构过程在数学建模上,可以归约到在其所能表达的构型空间中的两个构型之间找到一条满足某种条件的最优路径.本文围绕自重构模块化机器人的路径规划方法,从单个模块的运动方式、实现运动所需的模块硬件结构、多个模块整体间的组织方式、构型空间表示方法以及路径规划算法这5个层次,对国际上已有代表性工作进行总结分析,并在此基础上指出现有工作中存在的问题与挑战,提出可能的发展方向并对以后的工作作出展望,为研究人员系统性地了解自重构模块化机器人路径规划方法的现状以及开展后续研究提供了思路.Self-reconfigurable modular robots （SRMRs） are a special type of robots that can change their shapes and functions according to different tasks and environments. Such a robot is usually constructed using connected modules, each of which can encapsulate a simple function independently and also communicate with each other. Complex tasks can be completed by those connected modules collaboratively. In recent years, SRMRs have attracted considerable attention from both the academia and industry because of their versatility and flexibility. The path planning problem for the transformation of an SRMR is an important but not a well-solved problem, which can be considered as finding an optimal path in the configuration space where every point represents a feasible configuration of the SRMR. To provide a systematic overview of this research, we review the existing approaches considering five different aspects of SRMRs, including the type of motion on a single module, hardware for different motions, connectivity between modules, representation of a configuration space, and path planning algorithms. Aiming at motivating more research into SRMRs, the problems in existing approaches are analyzed and challenges in future work are summarized at the end of this paper.

关 键 词：模块化机器人 自重构 路径规划 NP完全问题 近似算法 计算几何 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]