Research On Design And Control Of Cable-Driven Manipulator For High-Payload Nursing Transport Tasks

Using intelligent robots to participate in nursing work is an effective means to solve the contradiction between the increasing number of elderly population and the serious shortage of nursing staff.The nursing robot should have a high load capacity,which can replace the nursing staff to help the elderly turn over,or carry the elderly between the hospital bed and wheelchair.However,the existing robots have the limitations of small load capacity or excessive self-weight,which cannot meet the actual needs of nursing.Therefore,based on the high load demand of the manipulator for nursing transport tasks,a cable-driven serial-parallel manipulator is developed in this dissertation.The cable-driven parallel mechanisms,high-load shoulder and elbow joints,rigid-flexible coupling wrist joint,kinematics modeling and optimization,dynamic modeling,and motion control method are studied theoretically and experimentally.A novel cable-driven serial-parallel high-load manipulator is proposed.The shoulder,elbow and wrist joints of the manipulator are all cable-driven parallel mechanisms.Combined with a special two-stage deceleration mechanism composed of harmonic reducer and pulley,the load capacity of the manipulator is greatly improved.Furthermore,the parameters of the shoulder and elbow joints are optimized to maximize its load capacity.Through structural optimization and the use of guide pulleys,the routing of the driving cables of the three joints is decoupled.In order to ensure the safety of the end effector of the manipulator in contact with human,a novel stiffness-adjustable rigid-flexible coupling mechanism based on a3-5R parallel mechanism is proposed,which is applied to the wrist joint of the manipulator.The kinematics of the cable-driven parallel mechanism and the spherical trajectory characteristics of the rigid-flexible coupling mechanism are analyzed by using screw theory.The variable stiffness working range of the rigid-flexible coupling mechanism is analyzed,and the rigid-flexible coupling characteristic equation is established.The stiffness and strength models of the joints are established,and the kinematics,strength and stiffness are simulated.Based on the joint rotation range,the workspace of the manipulator is analyzed and simulated.In order to solve the complex kinematics model of the proposed cable-driven serial-parallel manipulator efficiently,a parallel learning particle swarm optimizer(PLPSO)algorithm is proposed.Using the 28 test functions of CEC2013,the performance of PLPSO algorithm and other 9 reference PSO algorithms are evaluated.The Wilcoxon rank sum test and Friedman test are used to verify the superiority of the PLPSO algorithm.The structural bias of the PLPSO algorithm is studied,and the results show that the proposed random correction strategy and multi subgroup grouping can significantly improve the structural bias of the original PSO.Then,the PLPSO algorithm is applied to solve the kinematics of the proposed manipulator.The UR5 manipulator is also taken as a research object,PLPSO algorithm is compared with Differential Evolution algorithm(DE),Artificial Bee Colony algorithm(ABC)and Firefly algorithm(FA)for solving inverse kinematics.Finally,the PLPSO algorithm is compared with the numerical iterative method on UR5 manipulator and the proposed cable-driven serial-parallel manipulator to show the superiority of PLPSO algorithm.Due to the complex structure and cable routing of the cable-driven serial-parallel manipulator,its dynamic modeling is a challenging problem.A efficient dynamic modeling method is proposed.Firstly,the cable layout path of the proposed manipulator is clarified.Through the analysis of the cable-pulley unit model,the tension reduction equation of the cable-pulley unit is derived.The tension value of the driving cables of each joint of the manipulator is calculated based on the tension reduction equation.The contact force between the joints of the manipulator and the action equations of the rods are analyzed,the cable tension is optimized by the second norm minimization,and the Newton-Euler dynamic model considering the optimization of the cable tension and the friction of the cable-pulley is established.The flexible deformation of the joints of the cable-driven manipulator is further analyzed,and the rigid-flexible coupling dynamic model of the manipulator is established.Then,the Newton-Euler dynamic model of the manipulator is verified by simulation,and the rigid-flexible coupling factor of the manipulator is simulated,and the influence of the pulley reduction ratio and effective load on the dynamic coupling of the manipulator was analyzed.In order to realize the nursing task that the manipulator can lift the care recipient stably and keep the patient comfortable,a motion control method of the manipulator based on the optimal tactile force of the body being held is proposed.Based on the human body equivalent four-bar linkage dynamic equation,an evaluation model of human body tactile force is proposed.The constraint conditions of human joint angle and joint torque are given,and the PLPSO solution method for the tactile force model is proposed.The trajectory planning of the manipulator is carried out based on the holding posture parameters solved by the tactile force model,and a collaborative optimization method of the joint trajectory and the end trajectory of the manipulator is proposed.Combining the kinematics and dynamics equations,a dynamic feedforward control system is established,and the motion control of the manipulator based on the optimal tactile force is realized.Finally,the test platforms for the high-load shoulder-elbow joint manipulator and the rigid-flexible coupling wrist joint are established respectively,and the joint stiffness,workspace,speed and repeatability of the manipulator prototype are te sted,as well as the motion trajectory and stiffness of the rigid-flexible coupling wrist joint are tested.The prototype of proposed cable-driven serial-parallel manipulator is developed,the control system is established,and the experimental research platform of the manipulator is built.Through the motion control experiment of the manipulator,the real-time tension of the driving cables is tested to verify the kinematic and dynamic models.Through the end-tracking experiment of the manipulator,the trajectory accuracy of the end effector of the manipulator is tested and analyzed.Through the human-computer interaction experiment of the manipulator,the safety of the flexible wrist joint of the manipulator to human-computer contact is verified.The load experiment of the manipulator is carried out to test the load performance of the manipulator.The experiment of manipulator holding human body to verify the effectiveness of the proposed motion control method based on the comfort of holding posture.