Corridors optimization for robust MPC navigation in a cluttered environment

Abstract

This study addresses the issue of navigation in a cluttered environment of an autonomous vehicle subject to disturbances that may result from different sources, for instance: measurement errors, modeling errors, or state estimation errors. The main contributions of the paper reside in the use of a new definition of the concept of navigation corridor and an algorithm allowing to optimize these corridors while simplifying planning and ensuring the feasibility of the control problem based on a robust predictive control strategy.