Control Technology Of A Serial-parallel Flexible Manipulator

With the development of technology,robots are increasingly used in areas with high human-computer interaction requirements such as life services,medical assistance,and exoskeleton.Most of cooperative robots adopt serial-chain rigid structures,which have poor compatibility with people,thereby limit the application scenarios of robots to a certain extent.In this paper,based on the bionic mechanism of human arm muscle drive,a 6DOF serial-parallel manipulator(SoftArm-6)is designed,and the motion control algorithm is proposed.The verification experiments such as trajectory following,object grasping and human teaching are conducted.The main contents are as follows:Firstly,a 6-DOF serial-parallel compliant manipulator(SoftArm-6)is designed and built,which is composed of three serial degrees of freedom of the arm and three parallel degrees of freedom of the wrist.Flexible driving joints are formed by connecting motors in series with SEA.The kinematics model of the manipulator is established.Based on the principle of equivalent mechanism,the pose decoupling of the serial-parallel manipulator is realized.The forward and inverse kinematics solutions and velocity Jacobian matrix of SoftArm-6 are derived,and the effective workspace is solved.Secondly,the motion control method of SoftArm-6 is studied.A workspace trajectory planning method is proposed for the serial-parallel manipulator.Joint space real-time interpolation algorithm based on cubic polynomial and quintic polynomial variant is designed.The online trajectory teaching method based on human motion capture is studied using Kinect.Aiming to solve the problems of low position accuracy and vibration caused by the SEA flexible parts in the driving joint,a feedforward gravity compensation algorithm is designed based on the principle of virtual displacement.Through dynamic simulations of the serial-parallel manipulator using Adams software.It is proved that the algorithm can eliminate the influence of the gravity on joint position accuracy.Then,the control system of SoftArm-6 is designed and built.The control system includes the trajectory planning and state monitoring module of the upper computer and the position servo module of the lower computer.The hardwares of the control system are selected.The upper computer is an industrial computer and the lower computer uses STM32 microcontroller and real-time operating system Free RTOS.The system software is programmed and experimental platform is constructed.Finally,the performances of SoftArm-6 are verified through carrying out experiments including point-to-point positioning,linear trajectory following,gravity compensation and grasping operation.An online teaching experiment based on human motion capture is carried out,which verifies the following of SoftArm-6 to human motion and the effectiveness of online teaching method.