仿猿双臂手大阻尼欠驱动连续移动控制研究
发布时间：2018-11-06 21:22
【摘要】：随着仿生学的发展,人们对仿生悬臂摆荡运动等非线性欠驱动非完整约束系统控制的研究越来越多。猿猴等灵长类动物之所以能够在树枝间灵活地摆荡,得益于其通过长期进化而来的本能来控制自身的摆荡姿态,达到动能与势能之间灵活地转换。借鉴于猿猴的仿生摆荡控制原理,为了探索非线性欠驱动系统的控制的新领域,以及实现欠驱动机器人代替人类在高空桁架,外太空环境等非连续介质下连续移动的目的,本文研究了仿猿双臂手机器人在桁架杆下实现完全自主的大阻尼欠驱动连续移动抓杆的控制策略与方法。基于对灵长类生物摆荡运动机理的认识,本文首先介绍了本实验室所提出的在仿猿双臂手机器人大阻尼欠驱动抓杆策略。本文将连续移动抓杆从静止励振到下放停机细化为自启动励振阶段,调整构型大阻尼抓杆阶段,自运动调整松杆阶段,周期性摆荡连续移动阶段和下放停止阶段。对于欠驱动大阻尼阶段的机器人进行了运动学正逆解分析。对于欠驱动四杆机器人建立了动力学模型。结合机器人的运动学和动力学,本文对各个阶段的机器人进行控制模型的建立与控制策略的分析。针对目前仿猿摆荡机器人抓杆缺乏可靠性评估的现象,基于李雅普诺夫定理分析了机器人大阻尼的退转速度对于抓杆稳定性的影响,提出了速度相关和力矩相关的退转稳定性条件。针对三杆抓杆的可抓握范围不足的缺陷,本文提出了四杆协调反馈抓杆控制策略。为了解决实际机器人退转反馈延迟问题,提出了退转预估补偿策略。利用最优控制理论,对摆荡阶段的机器人进行了关节速度最小同时摆荡后重心最高的主驱动关节前馈轨迹的求解。根据几何关系,分析了自运动调整阶段,松杆阶段的需要满足的运动学约束条件。为了保证机器人各个阶段控制切换的稳定性,设计了各阶段间控制器切换条件,整合成一个总的能实现仿猿双臂手机器人完全自动化的连续移动控制器。建立ADAMS-Simulink联合仿真控制系统。对虚拟实验环境下的仿猿双臂手机器人分别进行了杆水平间距0.4m和0.6m的抓杆连续移动仿真。仿真结果表明所提出的方法能实现各个运动阶段间柔顺切换;四杆协调抓杆显著提高了大阻尼欠驱动反馈抓杆的可抓握范围,实现了机器人一次摆荡抓握目标杆的能力。仿真验证了所设计的控制器的有效性,以及所提出稳定性理论的正确性。
[Abstract]:With the development of bionics, more and more researches have been made on the control of nonlinear underactuated nonholonomic constrained systems such as the swing motion of bionic cantilever. Primates such as apes are able to swing flexibly between branches thanks to their long-evolved instincts to control their swinging posture and achieve a flexible conversion between kinetic energy and potential energy. In order to explore the new field of control of nonlinear underactuated system and to realize the continuous movement of human beings in discontinuous medium such as high altitude truss and outer space environment, this paper draws lessons from the theory of bionic swing control of apes, in order to explore the new field of control of nonlinear underactuated system. In this paper, the control strategy and method of the continuous moving grab rod with large damping underactuation under the truss bar for the ape-like dual-arm robot are studied. Based on the understanding of the mechanism of the swing motion of primate organisms, this paper first introduces the large damping underactuated grab rod strategy proposed in our laboratory. In this paper, the continuous moving grab bar from static excitation to down-down stop is divided into self-starting excitation stage, adjusting configuration with large damping grip stage, self-motion adjusting loosing stage, periodic swing continuous moving stage and down-down stopping stage. The kinematics inverse solution of the underactuated robot with large damping stage is analyzed. The dynamic model of underactuated four-bar robot is established. Combined with the kinematics and dynamics of the robot, the control model and control strategy of each stage of the robot are established and analyzed in this paper. In view of the lack of reliability evaluation of the grab rod of the ape-like swinging robot at present, based on Lyapunov theorem, the influence of the speed of the large damping of the robot on the stability of the grab rod is analyzed. The retrograde stability condition of velocity and torque correlation is proposed. Aiming at the deficiency of grasping range of three-bar grip, a four-bar coordinated feedback control strategy is proposed in this paper. In order to solve the problem of backward feedback delay of robot, a retrograde predictive compensation strategy is proposed. The optimal control theory is used to solve the feedforward trajectory of the main driving joint with the lowest joint velocity and the highest center of gravity after the joint swing. According to the geometric relations, the kinematics constraints of the self-motion adjustment stage and the loose pole stage are analyzed. In order to ensure the stability of the control switching in each stage of the robot, the switching conditions of the controller between the stages are designed and integrated into a continuous mobile controller, which can achieve the complete automation of the robot. The ADAMS-Simulink joint simulation control system is established. In the virtual experimental environment, the simulation of continuous movement of grasp rod with 0.4 m and 0.6 m horizontal distance is carried out respectively. The simulation results show that the proposed method can achieve compliance switching between different motion stages, and the four bar coordinated grip can significantly improve the gripping range of the large damping underactuated feedback grip, and realize the ability of the robot to grasp the target rod in one swing. Simulation results show the effectiveness of the proposed controller and the correctness of the proposed stability theory.
【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】：TP242
本文编号：2315501
[Abstract]:With the development of bionics, more and more researches have been made on the control of nonlinear underactuated nonholonomic constrained systems such as the swing motion of bionic cantilever. Primates such as apes are able to swing flexibly between branches thanks to their long-evolved instincts to control their swinging posture and achieve a flexible conversion between kinetic energy and potential energy. In order to explore the new field of control of nonlinear underactuated system and to realize the continuous movement of human beings in discontinuous medium such as high altitude truss and outer space environment, this paper draws lessons from the theory of bionic swing control of apes, in order to explore the new field of control of nonlinear underactuated system. In this paper, the control strategy and method of the continuous moving grab rod with large damping underactuation under the truss bar for the ape-like dual-arm robot are studied. Based on the understanding of the mechanism of the swing motion of primate organisms, this paper first introduces the large damping underactuated grab rod strategy proposed in our laboratory. In this paper, the continuous moving grab bar from static excitation to down-down stop is divided into self-starting excitation stage, adjusting configuration with large damping grip stage, self-motion adjusting loosing stage, periodic swing continuous moving stage and down-down stopping stage. The kinematics inverse solution of the underactuated robot with large damping stage is analyzed. The dynamic model of underactuated four-bar robot is established. Combined with the kinematics and dynamics of the robot, the control model and control strategy of each stage of the robot are established and analyzed in this paper. In view of the lack of reliability evaluation of the grab rod of the ape-like swinging robot at present, based on Lyapunov theorem, the influence of the speed of the large damping of the robot on the stability of the grab rod is analyzed. The retrograde stability condition of velocity and torque correlation is proposed. Aiming at the deficiency of grasping range of three-bar grip, a four-bar coordinated feedback control strategy is proposed in this paper. In order to solve the problem of backward feedback delay of robot, a retrograde predictive compensation strategy is proposed. The optimal control theory is used to solve the feedforward trajectory of the main driving joint with the lowest joint velocity and the highest center of gravity after the joint swing. According to the geometric relations, the kinematics constraints of the self-motion adjustment stage and the loose pole stage are analyzed. In order to ensure the stability of the control switching in each stage of the robot, the switching conditions of the controller between the stages are designed and integrated into a continuous mobile controller, which can achieve the complete automation of the robot. The ADAMS-Simulink joint simulation control system is established. In the virtual experimental environment, the simulation of continuous movement of grasp rod with 0.4 m and 0.6 m horizontal distance is carried out respectively. The simulation results show that the proposed method can achieve compliance switching between different motion stages, and the four bar coordinated grip can significantly improve the gripping range of the large damping underactuated feedback grip, and realize the ability of the robot to grasp the target rod in one swing. Simulation results show the effectiveness of the proposed controller and the correctness of the proposed stability theory.
【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】：TP242
【参考文献】
相关期刊论文 前10条
1 吴伟国;席宝时;;三自由度欠驱动机器人抓握目标杆运动控制[J];哈尔滨工业大学学报;2013年11期
2 董云云;刘治品;张兰兰;于征华;;欠驱动机器人控制策略综述[J];传感器世界;2012年12期
3 张安彩;赖旭芝;佘锦华;吴敏;;基于倒转方法的欠驱动Acrobot系统稳定控制[J];自动化学报;2012年08期
4 宋红生;王东署;;仿生机器人研究进展综述[J];机床与液压;2012年13期
5 孙宁;方勇纯;;一类欠驱动系统的控制方法综述[J];智能系统学报;2011年03期
6 张晓华;赵旖旎;程红太;;灵长类仿生机器人悬臂运动仿生控制综述[J];控制工程;2011年02期
7 任志全;余跃庆;周军;;水平运动的三自由度欠驱动机器人的位置控制[J];机器人;2010年06期
8 赵旖旎;程红太;张晓华;;基于能量的欠驱动双臂机器人悬摆动态伺服控制[J];西南交通大学学报;2009年03期
9 陈炜,余跃庆,张绪平;欠驱动机器人研究综述[J];机械设计与研究;2005年04期
10 许宏岩,付宜利,王树国,刘建国;仿生机器人的研究[J];机器人;2004年03期
相关硕士学位论文 前2条
1 顾夏东;仿猿双臂手机器人抓杆连续移动控制研究[D];哈尔滨工业大学;2015年
2 席宝时;三自由度仿猿双臂手欠驱动抓握桁架杆运动控制研究[D];哈尔滨工业大学;2013年
，本文编号：2315501
本文链接：https://www.wllwen.com/kejilunwen/zidonghuakongzhilunwen/2315501.html
