Optimization-Based Coordination of Automated and Human-Driven Vehicles in Confined Sites

Abstract

This article presents a two-stage optimization-based methodology to control the motion of automated vehicles (AVs) in confined sites where human-driven vehicles (HDVs) are also present. A centralized high-level controller computes the state and input trajectories of the AVs based on the estimated driving behavior of the HDVs such that occupancy conflicts are avoided in cross-intersections, narrow roads, and merge crossings. The high-level controller consists of a component tasked with computing the occupancy order of these zones and a subsequent component that utilizes the calculated order for obtaining the optimal state and input trajectories of the AVs. Through closed-loop simulations, we demonstrate the capability of the proposed approach to consider and partially control the HDVs and the benefit of the method to an alternative approach in terms of energy consumption.