A Local Motion Planner for Car-Like Robots in a Cluttered Environment

Abstract

This paper deals with the motion-planning problem for car-like robots (i.e., a mobile robot with nonholonomic and upper-bounded curvature constraints). In this paper, we propose an efficient planner for a simplified car-like robot model. Our motion planner approximates the derivative of configuration generated by a local holonomic motion planner that ignores motion constrainst, while guaranteeing collision avoidance. Using some simulations, we confirm the validity and efficiency of our algorithm.