To improve the compliance and stability of physical human-robot interaction

a method of physical human-robot interaction based on constrained optimization is developed and validated

aiming at the problems of impedance control methods that are easy to produce oscillations and need operator to apply additional forces.By designing and solving a constrained optimization problem

the robot motion reference velocity is obtained for a certain period after human-robot interaction

ensuring this velocity has the optimal human-robot interaction performance.The reference velocity of robot is further refined using an admittance model

enhancing the robot’s response to human-robot interaction forces to better conform to the natural contact relationship.In validation experiments conducted on a single-joint human-robot interaction platform

the method of constrained optimization significantly reduces robot jitter while responding to human body motion

improving the robot’s compliance on the premise of ensuring stability.