A Robust Impedance Controller for Improved Safety in Human–Robot Interaction

This paper presents a novel impedance controller modified with a switching strategy for the purpose of improving safety in human–robot interactions. Under normal operating conditions, an impedance controller is enabled when adequate tracking performance is maintained in the presence of bounded disturbances. However, if disturbances are greater than anticipated such that tracking performance is degraded, the proposed controller temporarily switches modes to a control strategy better apt to limit control inputs. With disturbances returning to the prescribed bounds, tracking performance will be restored and the impedance controller will resume for nominal operation. The control parameters are constrained by a few conditions necessary for smooth operation. First, a pair of equality constraints is required for the control signal to be continuous when switching control modes. Second, a Lyapunov analysis is performed to formulate an equality constraint on the control parameters to ensure only a single switch occurs when changing control modes to avert control chatter. Third, a matrix inequality constraint is necessary to ensure a robust positive invariant set is formed for when impedance control is active. Numerical simulations are provided to illustrate the controller and conditions. The simulation results successfully validate the presented theory, demonstrating how the constraints yield a continuous control signal, eliminate switching chatter, and permit robustness to disturbances.