Guaranteed safe navigation via state-constraints induced by feedback control

Abstract

A fast, novel, guaranteed collision-free, sensor-based navigation technique designed using an integrated approach towards planning and control for 2D unicycle mobile robots is proposed in this work. Unlike existing works on navigation that investigate planning to initiate the design process for integration, the proposed approach commences by finding state-constraining properties of feedback control for the design of a safe navigation controller using the sensor measurements. The invariant set describing the state constraints renders an incremental planning algorithm. As a consequence, the explicit geometric structure of the state constraints is directly utilized for progressing towards the target goal point while ensuring that the system’s trajectories are confined within the free regions of the workspace. The simulation and experimental results show that the proposed integrated control and planning technique not only renders a faster sensor-based navigator guaranteeing to reach the reachable target in an unknown environment but also inherently drives the robot towards maximum clearance path without explicitly planning for the same.