A Multi- Lane Cellular Automata Model Considering Connected Automated Vehicle Platoons for Mixed Traffic Flow

Vehicle-to-vehicle (V2V) and vehicle-toinfrastructure (V2I) technologies enable connected and automated vehicles (CAVs) to drive in platoons and realize cooperative driving. To investigate the impact of CAVs and platoons on multi-lane mixed traffic flow, this paper proposes a multi-lane cellular automata model considering the self-organized strategy of CAV platoons, which can describe the LC behaviors of CAV platoons. First, different car-following (CF) modes are introduced, and the CF characteristics and the lane-changing (LC) motivations of vehicles in different CF modes are analyzed. Then, based on CF characteristics and LC motivations, CF rules and LC rules for vehicles in different CF modes are designed, and the multi-lane cellular automata model considering the self-organized strategy of CAV platoons is developed. Finally, the impact of the LC behaviors of CAV platoons on the mixed traffic flow is investigated based on the numerical simulation. Numerical simulation results showed that: (1) benefiting from the LC behaviors of CAVs, CAVs can significantly improve traffic capacity and average velocity in the two-lane scenario compared to the single-lane scenario. When the penetration rate (PR) is 80%, the capacity of one lane and average velocity in the twolane scenario increases by 14.3% and 41.4%, respectively, compared to the single-lane scenario. (2) under the same PR of CAVs, the LC frequency showed a tendency of increasing and then decreasing with the density increase. Besides, as the PR of CAVs increases, the LC frequency decreases. (3) CAVs are significantly more effective in alleviating traffic congestion at a low density than at a high density. (4) when the PR of CAVs is not higher than 40%, the maximum platoon size has almost no effect on traffic capacity and velocity. As the PR of CAVs increases, the effect of maximum platoon size gradually becomes significant.