Travel time delay model with lane management policy for connected automated vehicles

This paper proposes a general framework of the travel time delay model with multi-lane road segments to consider different lane management policies (LMPs) for connected automated vehicles (CAVs). First, two LMPs for CAVs and a baseline lane policy are introduced. Second, the car-following modes and their proportions in CAVs dedicated lanes (CDLs) and normal lanes (NLs) with different LMPs are analyzed, respectively. Then, the multi-lane travel time delay model with different LMPs is derived based on the point queue (P-Q) model, and the relevant properties are proposed and proved. Finally, multiple parameters, such as the traffic demand, the penetration rate (PR) of CAVs, and the number of lanes and CDLs, are analyzed to assess their impact on the optimal split ratio, average reduced delay and optimal PR interval based on the proposed model, and the optimal number of CDLs and optimal LMP in different scenarios are given. Results show that (1) the optimal split ratio is always 0 until the PR reaches a critical value. Subsequently, the optimal split ratio increases with the PR; (2) only when the demand exceeds the critical demand and the PR is in the optimal PR interval, the (C, M) and (C, H) policies are possible to reduce the average delay; (3) compared with baseline lane policy, the optimal CDL settings can reduce the average delay by 100 % at maximum; (4) the optimal LMP is the baseline lane policy when the PR or the demand is low, while the optimal LMP is (C, M) policy when both the PR and the demand are high. This work provides insights into the impact of LMPs on travel time delay and helps decide the optimal number of CDLs and the optimal LMP.