Model predictive control of connected vehicles under automated driving at path-free signal-free intersections

Abstract

In this paper, we tackle the problem of real-time multi-vehicle interactions at signalfree and path-free intersections, which we call plazas, aiming for an efficient use of the intersection space. We propose a method that combines the path planning and path following tasks of connected vehicles under automated driving in one formulation. The method extends the standard nonlinear model predictive contouring control for application at intersections via the generation of multiple timeindependent reference paths. The goal is to maximize the progress of each vehicle along its reference path and penalize the rate of the control inputs while handling the continuous arrivals of the vehicles. The method is demonstrated using numerical simulation of vehicles at an intersection. The simulation results show that the proposed method achieves optimal and safe trajectories of the vehicles with driving flexibility and the possibility of better utilizing the intersection space.