面向抓取任务的移动机械臂运动规划研究

发布时间：2018-05-15 07:57

本文选题：移动服务机器人 + 移动机械臂　；参考：《哈尔滨工业大学》2017年硕士论文

【摘要】：随着机器人技术和社会经济的发展,移动服务机器人正在被国内外研究人员广泛的研究,而面向室内环境的移动机械臂因其良好的稳定性和环境适应性已成为研究热点。本文基于ROS(Robot Operating System),在Gazebo中搭建了用于室内环境中的移动机械臂系统,采用矢量法推导了其运动学方程,并提出了相应的自适应分层规划方法,在Gazebo仿真环境中进行了静态和动态两种目标抓取实验。针对室内环境中的抓取问题,结合移动机械臂实际的抓取需要,设计了一个自适应二指手。该自适应二指手采用了欠驱动的三关节手指结构,而手指的各关节可以转动,完成对目标的自适应抓取。同时为了更好地对手爪进行控制,建立了欠驱动手爪的动力学模型,对其动态静力学进行了分析,得到了输入力和接触力之间的关系。欠驱动手爪可以采用捏取和握取两种方式进行抓取,通过抓取不同形状的目标,验证其抓取的稳定性。完成了移动机械臂的系统搭建后,开展了移动机械臂的抓取规划研究。基于移动机械臂自身的特点,采用了矢量法对其运动学进行了分析,并建立了移动机械臂的运动学模型。考虑到移动机械臂运动的效率的问题,提出了自适应分层规划的方法,其中移动机械臂的规划划分为全局规划层和局部规划层,将移动平台的导航规划划分为全局规划和局部规划两部分,而机械臂的空间运动规划分为静态和动态分别进行分析,同时将提出的分层规划的算法在MATLAB中进行有效性验证。在上述工作的基础上,完成了移动机械臂在Gazebo中的移动平台的导航控制器和机械臂的空间运动控制器,以及其他控制参数的设定。采用了ROS实现对Gazebo下移动机械臂的运动控制,在仿真试验中,运用本文所述的自适应分层规划方法,成功的完成了移动机械臂静态和动态两种抓取。同时表明本文提出的机械臂和移动平台协同规划方法适用于室内环境中面向多目标的抓取任务。
[Abstract]:With the development of robot technology and social economy, mobile service robot has been widely studied by researchers at home and abroad, and mobile manipulator for indoor environment has become a research hotspot because of its good stability and environmental adaptability. Based on ROS(Robot Operating system, a mobile manipulator system for indoor environment is built in Gazebo. The kinematics equation is derived by vector method, and the corresponding adaptive hierarchical planning method is proposed. Static and dynamic target capture experiments are carried out in Gazebo simulation environment. An adaptive two-fingered hand is designed to solve the grab problem in indoor environment. The adaptive two-fingered hand adopts a three-joint finger structure with underactuation, and each joint of the finger can be rotated to achieve the adaptive grasp of the target. At the same time, in order to control the opponent's claw better, a dynamic model of underactuated claw is established, and its dynamic statics is analyzed, and the relationship between input force and contact force is obtained. The underactuated claws can be grasped by two ways, pinching and gripping, and the stability of grab can be verified by grasping different shapes of objects. After completing the system construction of the mobile manipulator, the grab planning research of the mobile manipulator is carried out. Based on the characteristics of mobile manipulator, the kinematics of mobile manipulator is analyzed by vector method, and the kinematics model of mobile manipulator is established. Considering the efficiency of mobile manipulator motion, an adaptive hierarchical planning method is proposed, in which the planning of mobile manipulator is divided into global planning layer and local planning layer. The navigation planning of mobile platform is divided into two parts: global planning and local planning, while the spatial motion planning of manipulator is divided into static and dynamic analysis, and the proposed hierarchical planning algorithm is validated in MATLAB. On the basis of the above work, the navigation controller of the mobile platform and the space motion controller of the manipulator in Gazebo are completed, as well as the setting of other control parameters. The motion control of mobile manipulator under Gazebo is realized by ROS. In the simulation experiment, the static and dynamic grasping of mobile manipulator is accomplished successfully by using the adaptive hierarchical planning method described in this paper. It also shows that the cooperative planning method of manipulator and mobile platform proposed in this paper is suitable for multi-target oriented capture tasks in indoor environment.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP241

【参考文献】