Optimal planning of dedicated lanes to maximize mixed highway traffic capacity

Abstract

With the rapid development of communication and automation technologies, connected and autonomous vehicles (CAVs) have gained increased attention as a potential solution to address the issue of traffic congestion. However, due to long transition period towards a fully autonomous vehicle fleet, it is crucial to understand the flow of mixed traffic that comprises both CAVs and human-driven vehicles (HDVs) and to find an efficient operation strategy. In this paper, we propose a tractable mathematical model for analyzing mixed traffic, considering diverse headways and CAV penetration rates. We determine the optimal CAV penetration rate for each lane through an optimization framework to maximize the capacity of multi-lane highways. We design two scenarios that reflect the level of CAV technology, and conduct numerical examples of multi-lane highways. Furthermore, we investigate the sensitivity of the optimal capacity and planning to changes in the headway between CAVs. The results reveal that it is optimal to sequentially assign CAVs from one lane to multiple lanes, keeping other lanes dedicated only to HDVs, as the total CAV penetration rate increases. Moreover, the presented framework enables to prevent the negative impact of traffic capacity from smaller CAV penetration rate under the conservative scenario. The optimal operation strategy of HDV- and CAV-dedicated lanes based on our findings will greatly contribute to improving the mixed traffic flow of multi-lane highways.