Scaling up the Optimal Safe Control of Connected and Automated Vehicles to a Traffic Network: A Hierarchical Framework of Modular Control Zones

Abstract

We consider the problem of scaling up optimal and safe controllers for Connected and Automated Vehicles (CAVs) from a single Control Zone (CZ) around a traffic conflict area to an entire network. The goal is to jointly minimize travel time and energy consumption for all CAVs, while providing speed-dependent safety guarantees within a CZ and satisfying velocity and acceleration constraints. A hierarchical modular CZ framework is developed consisting of a lower level where decentralized controllers are used that combine Optimal control and Control Barrier Functions (OCBF) and a higher level where a feedback flow controller is proposed to coordinate adjacent CZs. The flow controller is parameterized by a terminal velocity constraint that serves as the interface between CZs. Simulation results show that the proposed modular control zone framework outperforms a direct extension of the OCBF framework to multiple CZs without any flow control.