Experimental Implementation of a Trajectory Planner for Autonomous Driving

Abstract

Automation of road vehicles faced a disruptive growth in the last decades and it’s expected to deeply change many aspects of the society. In last years we assisted to a spread of autonomous driving pilot projects and proof of concepts by means of experimental prototypes both from companies and research groups. In this paper, a trajectory planner for autonomous driving is implemented and tested on a prototype vehicle. A full description of Model predictive control (MPC) formulation adopted is provided and implementation choices are motivated. A performance and reliability analysis is presented by means of a comparison between simulation and experimental results collected on a racetrack. Conclusions and future development directions are drawn.