中国科学院机构知识库网格
Chinese Academy of Sciences Institutional Repositories Grid
移动机械臂的自主抓取与协调控制研究

文献类型:学位论文

作者焦继乐
学位类别工学博士
答辩日期2014-05-27
授予单位中国科学院大学
授予地点中国科学院自动化研究所
导师谭民 ; 曹志强
关键词移动机械臂 嵌入式视觉 自主抓取 协调搬运 运动控制 Mobile manipulator Embedded vision Autonomous grasping Coordinated transportation Motion control
其他题名Research on Autonomous Grasping and Coordinated Control of the Mobile Manipulator
学位专业控制理论与控制工程
中文摘要移动机械臂由于同时具有移动功能和操作功能,已成为移动机器人领域的研究热点。无论对于结构化环境中的工业机器人,还是融入人类生活环境的服务机器人,移动机械臂的自主抓取和协调搬运都具有重要的研究价值和广阔的应用前景。本文针对移动机械臂的系统设计、自主抓取和协调搬运方法开展研究,论文的主要内容如下: 首先,介绍了移动机械臂的研究背景与研究意义,对移动机器人路径规划、机械臂的控制、机器人视觉系统及控制方法进行了综述,阐述了移动机械臂的研究现状,并对论文内容和结构做了介绍。 其次,设计并实现了移动机械臂系统,包括移动平台、机械臂的机构设计、控制系统的软硬件设计,并给出了多移动机械臂通讯的信息结构。 第三,开展了基于模糊趋近的移动机械臂目标趋近与抓取的研究,设计双输入单输出模糊控制器,将嵌入式视觉所获取图像信息的相关处理结果作为模糊控制器的输入,通过模糊控制器实现移动平台对目标的趋近。当移动机械臂距目标较近时,基于视觉信息进行位姿调整,直至目标在其可操作范围内,而后基于逆向运动学实施目标抓取。 第四,面向移动机械臂在障碍物环境下的抓取任务,提出一种移动机械臂在自身及环境约束下基于Bezier路径规划的自主抓取方法。根据移动机械臂的本体参数、当前位姿和目标信息,选取合适的控制点,进而规划出一系列可行的Bezier路径,并以路径长度和曲率为评价标准进行最优路径的选取,且在趋近过程中路径受环境影响时更新路径。伴随着移动机械臂对目标的趋近,机械臂的姿态实时调整以使得目标在摄像头的视野范围内,同时给出目标丢失的机械臂搜索策略。在移动机械臂较近距离的精确调整结束后,基于逆向运动学和图像匹配实施目标抓取。 第五,针对未知环境下多移动机械臂的协调搬运问题,提出一种基于设定搬运模式的协调搬运策略。根据多移动机械臂系统的几何布局,获取系统的初始信息,结合物体和系统的约束,定义了缺省模式、缩进模式和倾斜模式三种协调搬运模式。引入多尺度通道的概念,在多移动机械臂系统运动过程中,结合环境信息和任务需求,综合搬运模式、通道长度、通道方向与目标方向的夹角,以及环境障碍物的影响等指标进行综合评价,做出当前环境下的最优决策。 最后,对本文工作进行了总结,并指出了需要进一步开展的研究工作。
英文摘要The mobile manipulator receives more and more attentions in robotics with its mobility and manipulation. No matter for the industrial robot in structured environments, or the service robot in our daily life, the research on autonomous grasping and coordinated transportation of the mobile manipulator are of important value with potential applications. This thesis focuses on the system design of the mobile manipulator, and the approach on the autonomous grasping as well as coordinated transportation. The contents are as follows: Firstly, the research background and its significance of the mobile manipulator are given. The survey of path planning method of mobile robot, manipulator control, robot vision systems and their control methods is presented. The research development of mobile manipulator is then described. The contents and structure of this thesis are also introduced. Secondly, a mobile manipulator system is designed and developed, including the mechanical design of the mobile platform and manipulator, the hardware and software of control system. The information structure of communication among multiple mobile manipulators is also given. Thirdly, the research on fuzzy-based object approaching and grasping of mobile manipulator is carried out. A double input and single output fuzzy controller is designed for the approaching towards the object, and one of its inputs is the processing result of image acquired by an embedded camera. When the mobile manipulator is near to the object, it adjusts its pose based on visual information. Within the operation range of the manipulator, the object is grasped based on inverse kinematics. Fourthly, considering the grasping task in obstacle environments, a Bezier curve based path planning method is proposed, which considers the constraints of the mobile manipulator and obstacles. According to object information, the parameters and current pose of the mobile manipulator, the control points of Bezier curve are chosen. And then a series of feasible Bezier paths are planned, from which the optimal path is obtained on the basis of path length and curvature, and the path will be updated once the new detected obstacles block it. Along with the approaching of the mobile manipulator, the manipulator adjusts its posture in real-time to keep the object within the view of the camera. Also, a searching strategy is given when the object is lost. After a precise adjustment when near is done, the object is grasped based on inver...
语种中文
其他标识符201118014628009
源URL[http://ir.ia.ac.cn/handle/173211/6617]  
专题毕业生_博士学位论文
推荐引用方式
GB/T 7714
焦继乐. 移动机械臂的自主抓取与协调控制研究[D]. 中国科学院自动化研究所. 中国科学院大学. 2014.

入库方式: OAI收割

来源:自动化研究所

浏览0
下载0
收藏0
其他版本

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。