Passivity-Based Teleoperation With Interaction Force Constraints Using Hybrid Linear Model Predictive Control

Abstract

The bilateral teleoperation technique allows a human operator to interact with a remote environment and is generally applied in safety-critical scenarios. In such scenarios, even if the force feedback allows to feel the interaction with the environment, communication delay can lead to damages or side effects. For instance, in a critical scenario like surgery, different tissues tolerate different maximum forces and their manipulation often requires force control. Model Predictive Control (MPC) could be a solution to such situations thanks to its capability of embedding constraints in the optimization problem. In particular, linear MPC is usually employed in real-time control since it allows to have high control frequency. In this paper, we propose a hybrid-MPC based bilateral teleoperation. The hybrid system allows to explicitly handle the switching behaviour of the interaction force. Such methodology minimizes the number of switches between free motion and contact states. This reduced chattering provides smooth behaviour of the manipulator both in soft and hard contact scenarios.