Adaptive hybrid force-position control of a robotic manipulator

Abstract

A new hybrid force-position control method for uncertain robotic manipulator interacting with its environment is presented. First, the dynamical model system in the compliance frame is derives from the usual joint frame model and leads to two sets of equations due to the constraint associated to the contact surface. Next, the two dynamics are separately used for the synthesis of position and force tracking controllers. For the position control part, the design method consists of an estimated-parameters dependent coordinate transformation and a control law derived from a backstepping procedure. The force control law has two folds: first it compensates the dynamical interaction between the end-effector motion and the force induced by the environment and secondly, imposes a desired force using a proportional-like equation. Finally, a parameter-adaptation algorithm derives from stability criteria and dependent both on the position and force tracking errors. Simulation results on a four-degree of freedom robotic system tracking a triangle while maintaining a constant contact force prove the effectiveness of our solution. Key-Words : Control, bakstepping, force, position, adaptive.