Design and Experimental Validation of a Distributed Interaction Protocol for Connected Autonomous Vehicles at a Road Intersection

Abstract

This paper presents a fully distributed interaction protocol for connected self-driving cars negotiating the access a traffic junction. The vehicles coordination has been reformulated as an equivalent virtual platoon control problem. The desired vehicles spacing within the platoon is chosen so to avoid the side and rear-end collisions and is controlled by a cooperative non-linear control algorithm, based on potential functions. The asymptotic stability of the closed-loop system has been analytically proved by leveraging the LaSalle's Invariance Principle. The analysis of the convergence rate has resulted in an effective tuning tool ensuring that the desired formation is achieved before vehicles reach the intersection. Most notably, results from an in-vehicle experimental validation are presented. The experiments were carried out on three self-driving cars connected through a V2V communication infrastructure based on the IEEE 802.11p protocol. The experimental results confirm the theoretical analysis and reveal the effectiveness of the control approach for autonomously and safely negotiating a generic traffic junction.