Mobile Robot Navigation in a Cluttered Environment via Visual Predictive Control

Abstract

This paper presents a Visual Predictive Controller scheme for a differential drive robot navigating in a cluttered environment. We introduce an analytic model predicting the future state for this specific system Moreover, constraints guaranteeing the convergence of the control law, and avoiding occultations and collisions with obstacles are presented. A large set of results obtained in simulations highlights the interest and efficiency of the approach.