Comparative Analysis of Cooperative Driving Strategies for CAVs at On-Ramps

This paper focuses on cooperative driving strategies at on-ramps and comprehensively compares the performance of five representative strategies. The simulation results show that the dynamic programming (DP)-based, the grouping-based, and the rule-based strategies perform well in computation time and traffic efficiency, so all three strategies are recommended for practical use. We also show that improving traffic efficiency is usually at the expense of fairness, but a better trade-off between them can be realized in the modified DP-based and rule-based strategies. All cooperative driving strategies compared in this study will be integrated into CAVSim (a simulation platform dedicated to CAVs) for the convenience of researchers and the community.