Abstract
This paper presents a hierarchical data-driven/ model-based control architecture to robot implicit force control aiming at enhancing closed-loop performance. An inner data-driven controller, based on a recursive implementation of virtual reference feedback tuning, and relying on a model-based description of the robot compliance, prevents the deterioration of the force controller performance associated with environment modeling and identification. Then, an outer model predictive controller (MPC) acting as a reference governor selects on-line the optimal reference fed to the inner virtual reference feedback tuning closed-loop system in order to improve the closed-loop performance. The linear-time-invariant (LTI) formulation of the MPC allows to explicitly obtain the MPC control law and further demonstrate closed-loop stability. The effectiveness of the proposed control strategy is experimentally validated on force regulation experiments performed on different environment materials with a six-degree-of-freedom industrial robot equipped with a force/torque sensor.

This publication has 36 references indexed in Scilit: