An Adaptive Mobile Robot Control for Autonomous Following in Front of a Person

Abstract

This paper addresses the control of an autonomous mobile robot in a specific case, when the robot is following a person in front of him. Using a leader-follower formation approach combined with a look-ahead concept, a human-robot kinematic model in error coordinates with respect to a local Cartesian coordinate system is developed. A nonlinear feedback control is designed using local information from onboard sensor for the relative human-robot position and orientation. An adaptive control is proposed to estimate the unknown human linear and angular velocities. The effectiveness of the proposed robot control is validated via simulation tests and experiments.