中国科学院沈阳自动化研究所机构知识库
Advanced  
SIA OpenIR  > 机器人学研究室  > 学位论文
题名:
动态非结构环境冗余机械臂感知与规划
其他题名: Perception and Motion Planning of Redundant Manipulator in Dynamic Unstructured Environment
作者: 杜惠斌
导师: 刘光军
关键词: 动态非结构 ; 协作型机器人 ; 在线环境建模 ; 反应式避障规划 ; 运动学反解
页码: 130页
学位专业: 机械电子工程
学位类别: 博士
答辩日期: 2018-05-19
授予单位: 中国科学院沈阳自动化研究所
授予地点: 沈阳
作者部门: 机器人学研究室
摘要: 本文以国家自然科学基金资助项目“人机协作型新一代工业机器人基础研究”和中科院重点部署项目“新一代工业机器人关键技术开发”为依托,以提高协作型机器人的安全作业能力为目标,对冗余机械臂在动态非结构环境下的在线环境感知和在线轨迹规划问题展开研究,主要内容如下:协作型机器人在线环境建模与感知问题研究。协作型机器人能够在线感知工人等动态随机障碍物,是其避免碰撞、保证人与机器人以及环境安全的基础。本文用深度相机作为环境感知的外部传感器,研究协作型机器人环境模型建立及障碍信息获取方法。针对建模问题,采用线性不变性和旋量理论,研究了深度相机的快速部署手眼标定方法,并通过对深度图像特点的分析,建立了三维栅格离线映射模型,研究了环境信息在线更新与多传感器数据融合方法,实现了传感器盲区和被遮挡视野的快速地处理;针对障碍物在线感知问题,研究了从环境模型中分割机器人与障碍物并实时提取二者之间的最近距离和最近点的方法,为后续的在线避障规划提供支撑。协作型机器人在线轨迹规划问题研究。安全笼的取消使得机器人的工作空间由传统的静态已知结构化变成了动态随机非结构化环境。当机器人感知到工作空间内障碍物有可能干扰到其正常运行时,需要在线规划运动轨迹实现动态避障,以避免潜在的碰撞危险,同时要考虑作业任务的执行。本文采用基于动态系统在线轨迹生成的方法,研究实现了冗余机械臂在执行作业任务时的末端执行器动态避障。而针对机械臂本体连杆的避障问题,采用将避障任务转化为最优化问题约束条件的方法,以优先级/权重在线调节的方式实现了轨迹的在线规划。这种方法可以使机械臂在冗余能力范围内,按不影响其末端点作业任务的轨迹完成连杆避障,而当冗余能力不足时,平稳切换作业和避障任务的优先级顺序,从而确保不发生碰撞。协作型机器人运动学实时优化反解问题研究。自由度冗余设计目前被多数协作型机器人所普遍采用,其优势在于其对复杂的人机共融环境与任务具有更高的适应性。如何充分利用机械臂的冗余空间并满足机器人关节的限位、限速等运动约束,同时要满足与控制周期同步的实时性要求,是协作机器人在线优化反解的技术难点。本文采用线性规划方法展开研究,并将保证运动约束和实时求解设为首要目标,而将运动性能的优化放在次要位置,使冗余机械臂的优化反解适用于基于传感器数据的动态避障等在线应用,同时避免了因运动约束失效或无法及时获得关节运动指令造成机器人运动轨迹不可预知的潜在危险。本文进一步地对以上三个方面研究成果进行了仿真和实验验证,并搭建了仿真与实验的软硬件系统平台。仿真和实验结果都验证了本文提出的技术与算法的有效性和实用性。
英文摘要: The research in this dissertation is sponsored by the project of Basic Research Program of Human-Robot Collaboration from National Natural Science Foundation and the project of Key Technology Development of Next Generation Industrial Robot from Chinese Academy of Sciences. In this dissertation, the research on the safety assurance of collaborative robot in the dynamic unstructured environment is carried out, aiming at improving the environment perception and motion planning capacity of the new generation redundant manipulator. The main contents of this dissertation are as follows: Online modeling of the dynamic unstructured environment of collaborative robot is studied. The ability of collaborative robot online perceiving dynamic obstacles such as workers is the prerequisite of collision avoidance and safety guarantee for human, robot and environment. This dissertation uses the depth sensors like Kinect as the external sensors for environment perception to study the environment modeling and the obstacle information extracting. As for the environment modeling, a quick-deploy method of hand-eye calibration for depth cameras and an off-line mapping method of the depth image and 3-D grids for online updating and multisensor data fusion of environment model are proposed. These algorithms can deal with the blind or obstacle occlusion area of depth sensors quickly and efficiently. As for the obstacle information extracting, a method for filtering out the robot itself from the model and extracting the distance from obstacles to robot is proposed, providing support for subsequent online obstacle avoidance motion planning. Online motion planning of collaborative robot is studied. The concept of cage-free turns the traditional static structured environment of the manipulator into the dynamic unstructured environment. When the obstacles in the working space of the collaborative robot may disturb or collide with the robot, the robot should be able to change its trajectory to avoid the dynamic obstacles. In this dissertation, an online trajectory generation method based on dynamic system for the task of redundant manipulator is introduced. This method can drive the end-effector of the robot away from its path in order to avoid the dynamic obstacles. The dissertation proposed a method that transform the obstacle avoidance trajectories into the constraints of an optimized framework. Using this method the manipulator can avoid the obstacles as far as possible within the redundancy ability without affecting the end-effector. In addition, if the redundancy is insufficient, the technique of online priority/weight adjustment are used to deal with the obstacle avoidance of the multi points on the links of robot. Real-time inverse kinematics optimization of redundant manipulator. The Cartesian space trajectory obtained from reactive online motion planning is transformed back to the configuration space of the robot where the actuators are actually controlled. Redundancy is one of the characteristics of collaborative robots, which increase the ability of obstacle avoidance and the difficulty of inverse kinemics. Optimization is usually required to benefit the advantages of the redundant DOFs while avoiding the constraints on each of the joint actuators, i.e., the angular position limit, maximum velocity and acceleration. However, the optimization in real time presents challenges due to the computation complexity of the inverse kinematics under the actuator constraints. In this dissertation, the linear programming method is used for addressing the real-time optimized inverse kinematics of redundant robots with joint physical constraints. This method makes the optimized inverse kinematics of redundant robots suitable for the online application, such as dynamic obstacle avoidance. The dissertation has further tested the proposed algorithms in simulations and experiments on the testing system involving KUKA IIWA robot and Kinect sensors. The results have verified the electiveness and practicability of the proposed algorithms.
语种: 中文
产权排序: 1
内容类型: 学位论文
URI标识: http://ir.sia.cn/handle/173321/21779
Appears in Collections:机器人学研究室_学位论文

Files in This Item:
File Name/ File Size Content Type Version Access License
动态非结构环境冗余机械臂感知与规划.pdf(5410KB)学位论文--限制开放 联系获取全文

Recommended Citation:
杜惠斌. 动态非结构环境冗余机械臂感知与规划[D]. 沈阳. 中国科学院沈阳自动化研究所. 2018.
Service
Recommend this item
Sava as my favorate item
Show this item's statistics
Export Endnote File
Google Scholar
Similar articles in Google Scholar
[杜惠斌]'s Articles
CSDL cross search
Similar articles in CSDL Cross Search
[杜惠斌]‘s Articles
Related Copyright Policies
Null
Social Bookmarking
Add to CiteULike Add to Connotea Add to Del.icio.us Add to Digg Add to Reddit
所有评论 (0)
暂无评论
 
评注功能仅针对注册用户开放,请您登录
您对该条目有什么异议,请填写以下表单,管理员会尽快联系您。
内 容:
Email:  *
单位:
验证码:   刷新
您在IR的使用过程中有什么好的想法或者建议可以反馈给我们。
标 题:
 *
内 容:
Email:  *
验证码:   刷新

Items in IR are protected by copyright, with all rights reserved, unless otherwise indicated.

 

 

Valid XHTML 1.0!
Copyright © 2007-2018  中国科学院沈阳自动化研究所 - Feedback
Powered by CSpace