Merging sections on highways are identified as traffic bottlenecks, leading to congestion and accidents. The emergence of Connected and Autonomous Vehicles (CAVs) technology promises an optimized solution to on-ramp merging issues. Existing cooperative merging strategies typically focus on determining the merging sequence of vehicles in specific merging areas, overlooking the influence of the macroscopic traffic flow conditions on the merging process. In this paper, we innovatively introduce the Dynamic Cooperative Merging Assistance (DCoMA) strategy, a traffic management approach designed to enhance merging operations under variable traffic demands. At the macroscopic level, DCoMA employs the fundamental diagram of traffic flow to develop a platoon formation algorithm for mainline vehicles, tailored to the dynamic macroscopic states of traffic. The algorithm adaptively adjusts the size of platoons based on the volumes of both the mainline and the on-ramp. Subsequently, the spatio-temporal dynamics of these platoons function as the ’red phase’ of traffic signals, with the intervals between platoons analogous to the ’green phase’. This information is then converted into a time-series format and transmitted to all vehicles on the on-ramp. At the microscopic level, vehicles on the on-ramp alter their driving strategies in response to this time-series data, ensuring a seamless merging into the mainline flow without causing halting. Through simulations and comparative analysis with three existing strategies (i.e., CoopMA, ALINEA, and X-ALINEAQ), the proposed DCoMA strategy exhibits considerable potential for application across diverse traffic volumes and CAV penetration rates. The results indicate that the proposed approach can significantly improve the efficiency of the mainline and on-ramp, achieving a maximum improvement of 18.32%. More importantly, the DCoMA strategy effectively enlarges merging gaps and, coupled with the speed control capabilities of CAVs, significantly mitigates the risk of accidents and reduces exhaust emissions.
Read full abstract