Path Planning using a Dynamic Vehicle Model

This paper addresses the problem of path planning using a dynamic vehicle model. Previous works which include a basic kinematic model generate paths that are only realistic at very low speed. By considering higher vehicle speed during navigation, the vehicle can significantly deviate from the planned trajectory. Consequently, the planned path becomes unusable for the mission achievement. So, to bridge a gap between planning and navigation, we propose a realistic path planner based on a dynamic vehicle model

1 Introduction

Moving an autonomous vehicle is often divided in two phases. In the first one, a feasible path between two configurations is computed. Then, this path is followed by the vehicle, using the trajectory returned by the planner and a control law. Most of research works considered these two steps as independent and only focus on one of them. Unfortunately, the planning phase is strictly dependent on the model used during the navigation process.