Human Features-Based Variable Admittance Control for Improving HRI and Performance in Power-Assisted Heavy Object Manipulation

Abstract

In the first step, a 1-DOF power assist robotic system (PARS) was developed for object manipulation. The dynamics for human-robot co-manipulation of objects was derived that included human's weight perception. An admittance control scheme with position feedback and velocity controller was derived using the weight-perception-based dynamics. Human subjects lifted objects with the PARS, and human-robot interactions (HRI) and system characteristics were analyzed. HRI was expressed in terms of physical HRI (maneuverability, motion, safety, stability, and naturalness) and cognitive HRI (trust and workload), and performance was expressed in terms of manipulation efficiency and precision. A constrained optimization algorithm was used to determine the optimum HRI and performance. Results showed that inclusion of weight perception in the dynamics and control was effective to produce optimum HRI and performance for a set of hard constraints. In the second step, a novel variable admittance control algorithm was proposed, which enhanced the physical HRI, trust, precision and efficiency by 34.63%, 31.86%, 3.85% and 4.92% respectively, and reduced cognitive workload by 35.36%, and helped achieve optimum HRI and performance for a set of soft constraints. Effectiveness of the control algorithm was justified using a multi-DOF PARS for manipulating heavy objects.