【作者】 王东利；

【导师】 杜永贵；

【作者基本信息】 太原理工大学 ， 控制理论与控制工程， 2012， 硕士

【摘要】 在近几年来,空间机器人技术得到了各国航天部门的高度关注,太空的恶劣环境和昂贵的载人航天费用使得发展空间机器人技术很有必要性,本课题来源于总装国家航天921项目“面向空间站科学实验的遥操作概念研究”,本文研究工作紧密结合该项目,主要研究了空间站舱内冗余机器人模拟系统的总体设计和系统集成,重点研究了冗余机器人的运动学、轨迹规划和控制系统搭建。实验验证了该模拟系统在大时延的情况下,地面操作员能够稳定和精确地操作空间舱内机器人。本文的主要研究内容有如下几个方面：(1)首先综述了空间站舱内机器人实验平台的研究背景和国内外的研究现状,系统分析了舱内机器人的关键技术,包括冗余机器人的运动学分析和模块化机器人的构建。(2)对冗余机器人建立运动学模型,本系统冗余机器人由移动式导轨、七自由度模块化关节机器人和灵巧手组成。采用D-H法对冗余机器人建立正运动学模型,通过MATLAB仿真软件对正运动学方程进行验证。逆运动学的解法包括解析法和广义扩展雅克比矩阵法。(3)阐述了冗余机器人的整体控制算法,包括运动学逆解、轨迹规划算法和避障算法研究。由于解析法运算量比较大,不适合做实时控制,只能通过广义扩展雅克比矩阵法对其运动学求逆解,本系统机器人是冗余机器人,其雅克比矩阵不可逆,引入雅克比矩阵的伪逆,通过梯度投影法对伪逆解优化。用解析法求得的公式对结果验证,可有效避开机器人的奇异状态。轨迹规划算法采用定时插补法,在轨迹规划中加入避障算法,完成对冗余机器人的稳定安全控制。(4)研究了空间舱内冗余机器人系统的总体设计方案和硬件构成,包括导轨控制系统、七自由度模块化关节机器人系统、手控器系统、灵巧手系统、视觉系统和空间站模拟系统。(5)分析研究了空间舱内机器人系统的整体算法设计和结构组成,主要包括手控器控制算法、灵巧手控制算法、冗余机器人控制算法、主动视觉控制算法、Kinect视觉控制算法和虚拟现实技术。上位机负责监视和控制机器人系统,下位机负责接收上位机控制命令,向设备发送控制命令和向上位机返回数据信息,包括力反馈和视觉信息。最后,对已开展的工作进行总结,并对今后开展的工作进行了展望。更多还原

【Abstract】 In recent years, space robot technology gets the high degree of attention from space agencies all over the world. The hostile environment of space and the expensive expenses of manned spaceflight make it essential to develop the space robot technology. This work is supported by Project921Human Spaceflight Programme, which named "The tele-robot concept research for space station scientific experiment". This paper which is tightly integrated with the project, studies the overall project design and the system integration of the robot simulation system in cabin space station, especially, the kinematic, trajectory planning and controls for redundancy robot. The experimental results prove the high efficiency of this simulation system, and with a large delay existed, a ground operator could still manipulate the space robot steady and accurately.The paper consists of the following parts:This paper describes the research background and the status of research at home and abroad, and makes a systematic analysis about key technologies for capsule robot, including the kinematics analysis for redundancy robot and the structure for modular robot. It makes kinematics models for redundancy robot, which is composed of mobile guide, modular robot with seven degree of freedom joints and dexterous hand. Direct kinematics model is made for the robot using D-H method, and then MATLAB simulation software is used for checking. The inverse kinematics method consists of analytical solution and method extended from Jacobin matrix.The integral control algorithm for redundancy robot is expounded in this paper, including research of several algorithms such as inverse kinematics solution, trajectory planning and object detection algorithm. Due to the large computation quantity of analytical solution, it does not apply to control the robot in real time, and as a result, it could only use Jacobin matrix to solve inverse kinematics. However, the robot in this system is a redundancy robot whose Jacobin matrix is irreversible, and then the gradient projection method is used to optimize the inverse kinematics. Results could be confirmed by formulas calculated by analytical solution, effectively avoiding robot’s singular state. Trajectory planning method adopts timed interpolation method. Joint with object detection algorithm, it achieves a security and stability control system for redundancy robot.The overall project design and hardware structure of redundancy robot system is discussed in this paper, including guide control system, modular robot system with seven degree of freedom joints, manual controller system, dexterous hand system, vision system and space station simulation system. The overall algorithm design and structure components of the robot system is analyzed and researched, including manual controller control algorithm, dexterous hand control algorithm, redundancy robot control algorithm, active visual control algorithm, Kinect visual control algorithm and virtual reality technique. The upper computer is responsible for monitoring and controlling the robot system, while the lower computer is for receiving commands from the upper, sending control commands to device, and returning data information (such as Force Feedback information and visual information) to the upper.Finally, there is a conclusion made for the existing work and a prospect for the future work.更多还原