The Human Upper Arm Endpoint Stiffness Measurement,Estimation And Human Robot Interaction

Impedance control is the key technology to improve the flexibility of human-robot interaction.Stiffness parameter is an important parameter in impedance and plays a decisive role in robot compliance control domain.It is helpful to realize the robot's anthropomorphic stiffness with the identification of human stiffness parameter learned from that the mechanism of human body stiffness to resist environment disturbances,which can improve the compliance and anthropomorphism of the interactive robot.The surface electromyogram signal(sEMG)contains abundant biological information that generates before human body movement and can be detected non-invasively,so it is suitable as a human body stiffness information decoding signal and to send stiffness adjustment command to the robot.In this thesis,the human upper arm endpoint stiffness is measured with the measuring device developed domestically and estimated real-time based on sEMG.There four main parts in this dissertation,including:(1)The establishment of the human upper arm endpoint stiffness measurement platform.The stiffness measurement platform designed domestically can be divided into three modules according to function,comprising measurement module,transmission module and drive module.Among them,the key mechanism includes a cam mechanism and a planetary gear mechanism,and the horizontal random vibration is achieved by the rotational speed ratio between the cam and planetary gear.(2)Measurement of the stiffness of the upper arm endpoint based on small perturbation theory.The measurement experiment was carried out on stiffness measurement device,and then to analyze the precision of the stiffness of the spring-dynamics model and stiffnessdamping-inertial dynamic model and graph the stiffness characteristics with the ellipse.(3)Human upper arm stiffness estimation model based on sEMG.The joint torque estimation model parameters were identified with the sEMG on eight muscles of upper arm and joint torque to calculate the index of co-contraction around joint(IMCJ).And then analyze the influence of stiffness by restoring force and IMCJ.Finally establish the human upper arm endpoint stiffness based on IMCJ.(4)Development of admittance control system based on robot operating system(ROS).The kinematic coordinate system of the manipulator is established to realize the inverse kinematics solution.The admittance controller based on the force feedback of the manipulator arm is designed and tested by different impedance parameters to change and explore the control performance and the feasibility which applies the proposed stiffness estimation method to human-robot interaction domain.