Cooperative Merging in Mixed Traffic: A Mobile-Edge Hybrid Control Framework

Abstract

This paper addresses decision-making challenges in mixed traffic environments comprising both conventional human-operated vehicles (HVs) and connected automated vehicles (CAVs). Our proposed framework is exemplified using a ramp merging scenario and is structured as an optimization problem, in which a merge sequencing problem and a trajectory planning problem are embedded and solved by a bi-level hybrid centralized-decentralized model predictive control (HMPC) approach. The HMPC framework we introduce leverages centralized edge computing for efficient merge decision optimization through a dynamic-programming approach and decentralized mobile computing for distributed trajectory planning through three different optimization algorithms. Simulation results show that compared to open-loop control, the proposed framework can ensure system efficient ramp-merging control, and exhibits robustness in the presence of uncertainty caused by the stochastic driving behaviors of HVs. In addition, it is found that mobile-edge hybrid framework can reduce the computational time to the millisecond-level, potentially meeting real-time computational requirements.