Research On Robot Force Control Based On Adaptive Admittance Framework

At present,people mainly rely on robots with high repetitive positioning accuracy and load capacity to perform some repetitive and simple tasks instead of people.With the continuous development of robot technology in recent years,a small cooperative robot specially designed for human-machine cooperation has been widely used.As a result,higher requirements are put forward for the control algorithm of the robot.Generally,cooperative robots are required to have active compliant control algorithm to better perform human-computer interaction tasks and force tracking tasks.Therefore,this paper mainly studies the cooperative robot manpower control algorithm based on adaptive admittance control framework.First,the modify DH method is used to model the Franka panda robot and derive its Jacobian matrix and the solution process of forward and inverse kinematics of the robot,laying a theoretical foundation for subsequent experiments and analysis.General admittance control uses a six-dimensional force sensor installed in the robot TCP as the input of the force signal,which usually couples a lot of unwanted force information,so it is essential for the algorithmic compensation of the end effector.In this paper,the correctness and feasibility of the gravity compensation algorithm of the six-dimensional force sensor are verified by mathematical derivation and experiments.Then,since the admittance algorithm of robot is mainly used in humancomputer interaction in free space,the admittance parameter selection and control strategy of variable admittance control based on human operation intention are discussed.Since the whole control system forms a closed loop during the interaction between human and robot,based on modeling the impedance of human upper limb,the absolute stability theorem is used to analyze the stability boundary to determine the selection of admittance parameters.At the same time,a security strategy based on admittance control framework is proposed to solve the problem of singularity in dragging teaching and the limitation of working space of cooperative robot.Finally,the force tracking problem of robot in constrained environment is discussed.In the process of force tracking between the robot and the unknown environment,the interaction force between the robot and the environment often deviates from the expected force due to the unknown environment information.In order to eliminate the steady-state error of force,an adaptive method to the environment is used to estimate the position and stiffness of the environment in real time.To solve the problem of slow manpower control in some situations,nonlinear feedback and force outer loop PID were added to solve the problem.The proposed algorithm was simulated and verified by simulink,and the selection of admittance parameters was determined in the simulation and experiment,and the selection of control parameters was optimized based on the force control response index.