Intersection Crossing in Mixed Traffic Flow Environment leveraging V2X information

Abstract

The paper addresses the problem of safe crossing a non-signalized intersection in a mixed traffic scenario composed of both connected human-driven and autonomous vehicles. To solve this problem, we propose a control strategy that leverages information shared via V2X for extending the line-of-sight of the on-board ranging sensors. The control protocol augments the classical I-ADAS approach based on local measurements with an additional networked action that has the aim of preventing oncoming colliding situations. The control strategy is purposely designed for counteracting the unavoidable presence of communication impairments that turns into multiple time-varying delays affecting the information sent via the cyber-space. The numerical analysis, carried out leveraging the Matlab/Simulink platform, discloses the effectiveness of the approach for a challenging case where vehicles dynamic conditions are such that collisions would have occurred at the intersection if the vehicle would only be equipped with range sensors.