【作者】 彭涛；

【导师】 蔡鹤皋；

【作者基本信息】 哈尔滨工业大学 ， 机械工程（专业学位）， 2018， 硕士

【摘要】 相对于传统工业机器人,人机协作型机器人具有良好的自主性和柔性操作性能,能够与人共融、协作完成复杂任务。面向柔性操作任务,单纯的轨迹示教不能够满足阻抗变化的需求,本课题以人机协作技能传授为背景,基于人手臂sEMG信息建立手臂末端阻抗模型,并且依据该模型将人对人机协作过程中阻抗变化的判断映射到机械臂的变阻抗控制模型,从而实现轨迹和阻抗参数共同示教,提高人机协作完成复杂柔性作业任务的性能。首先,对iiwa人机协作机器人的运动学进行分析。采用D-H法建立机器人连杆坐标系,根据坐标转换矩阵推导机器人正运动学方程;并且,采用微分变换法推导了机器人雅克比矩阵,采用逆雅克比矩阵实现笛卡尔速度向关节速度转换,从而完成机器人逆运动学解算;利用MATLAB Robotics Toolbox工具箱建立虚拟仿真环境进行运动学仿真验证。并对iiwa机器人工作空间进行评估,为人机协作过程中可操作空间提供理论依据。其次,对基于sEMG人体上肢末端阻抗估计进行研究。建立上肢末端刚度模型,分析上肢末端刚度大小、方向等属性。在表面肌电信号的采集及预处理基础上,建立基于sEMG的上肢末端刚度估计模型;通过实验对上肢末端力映射T_F和力映射零空间肌电信号向上肢末端刚度映射ψ两个模型参数进行辨识,从而建立基于sEMG信息的手臂末端刚度模型。再次,对基于sEMG的变阻抗控制展开研究。建立基于sEMG的变阻抗控制架构以及参数映射准则;根据人-机阻抗物理与数学模型,仿真分析阻抗参数对控制性能的影响;对力信号进行重力补偿与滤波算法研究;进行力的随动性能验证实验与基于sEMG变阻抗控制器性能验证实验,为下一步基于sEMG的人机协作柔性示教实验奠定基础。最后,搭建人机协作变阻抗实验平台,包括硬件系统和软件系统搭建;并且设计复杂轨迹跟踪实验和柔性作业复现实验两个典型实验。验证本文提出的基于sEMG信息的人机协作机械臂变阻抗示教的可行性,并且对其性能进行评估。更多还原

【Abstract】 Compared with traditional industrial robots,the human-robot cooperative robot has good autonomy and flexible operating performance,and can complete complex tasks with human.For the flexible operation task,the simple trajectory teaching can not meet the demand of impedance change.In this subject,the end-point impedance model of the human arm is built on the basis of human arm s EMG information.Based on this model,the estimation of the impedance changes during the human-robot cooperation process is mapped to the variable impedance gain of the manipulator’s impedance controller.In order to improve the performance of complex and flexible tasks,human-robot cooperation is achieved through the joint control of trajectory and impedance parameters which learning from human tutor.Firstly,kinematics analysis of iiwa human-machine cooperation robot is carried out.The D-H method is used to establish the coordinate system of the robot links,and the forward kinematics equation of the robot is derived according to the coordinate transformation matrix,and the Jacobi matrix is derived by differential transformation.The inverse Jacobian matrix is used to transform the velocity of Cartesian space into the joint space,thus the inverse kinematics of the robot is calculated.The MATLAB Robotic toolbox is used to establish the virtual simulation environment where can carry out the kinematics simulation verification.The workspace of iiwa robot is evaluated to provide a theoretical basis for the workable space in the process of human-robot cooperation.Secondly,the impedance estimation of human upper limb based on s EMG is studied.The stiffness model of the upper limb was established to analyze the stiffness and orientation of the upper limb.On the basis of the collection and preprocessing of surface EMG signal,the stiffness estimation model of the upper limb based on s EMG is established.Through the experiment,two model parameters of the end force mapping T_F and the force mapped zero space EMG signal are mapped to the end stiffness of the upper extremity to establish the arm terminal stiffness model based on the s EMG information.Thirdly,the research of variable impedance control based on s EMG is carried out.The variable impedance control architecture based on s EMG and the parameter mapping criterion are established.According to the physical and mathematical models of human-robot interaction impedance,the influence of impedance parameters for the control performance is simulated and analyzed;the research of gravity compensation and filtering algorithm based on force signal was finished;theperformance of force servo and the s EMG based variable impedance controller were verificated experimental.The experiment lays the foundation for the next step of human-robot cooperation flexible teaching experiment on the basis of s EMG signal.Finally,the variable impedance human-robot cooperation experimental platform is built,including the hardware and software systems.Two typical experiments are designed for the complex trajectory tracking experiment and the flexible operation reiteration experiment.The feasibility of the variable impedance human-robot cooperation arm impedance teaching based on s EMG information propose here is verified and its performance is evaluated.更多还原