Cell transmission model of mixed traffic flow of manual-automated driving

In order to analyze the impacts of automated driving vehicles on the macroscopic traffic flow characteristics, the mixed traffic flow with manual driving vehicles and automated driving vehicles was considered as the study objective, and the cell transmission model (CTM) of mixed traffic flow under different proportions of automated driving vehicles was proposed. The car-following model proposed by Newell was used for the car-following model of manual driving vehicles, while the model calibrated by PATH program used the real vehicle experiments was employed for the car-following model of automated driving vehicles. The function relation of equilibrium space headway-speed was calculated according to the car-following models of manual and automated driving vehicles. The fundamental diagram model of mixed traffic flow was derived under different proportions of automated driving vehicles. In addition, the characteristic quantities such as the maximum capacity, the maximum jam density, and backward wave speed were calculated for the mixed traffic flow under different proportions of automated driving vehicles. Based on the CTM theory of homogenous traffic flow, the CTM of mixed traffic flow was proposed under different proportions of automated driving vehicles. The moving bottleneck problem was selected for example analysis, the influence times of moving bottleneck under different proportions of automated driving vehicles were calculated by using the mixed traffic flow CTM. The car-following models were used for the microcosmic numerical simulation on the moving bottleneck problem. The errors between the calculation results of the mixed traffic flow CTM and the microcosmic simulation results of car-following models were analyzed. The accuracy of mixed traffic flow CTM was validated. Research result shows that the proposed mixed traffic flow CTM can effectively calculate the influence time of moving bottleneck. Under different proportions of automated driving vehicles, the errors between the calculation results of the mixed traffic flow CTM and the microcosmic simulation results of car-following models are all below 52 s, and the relative errors are all below 10%, which indicates the accuracy of the proposed mixed traffic flow CTM in actual application. The mixed traffic flow CTM reflects the study idea from microcosmic to macroscopic. There are relationships between the microcosmic car-following models and the small-scale automated driving vehicle experiments being gradually implemented. The mixed traffic flow CTM can truthfully reflect the evolutionary process of mixed traffic flow on single lane in the background of automated driving under different proportions in the future, which enhances the application value of the model research. © 2020, Editorial Department of Journal of Traffic and Transportation Engineering. All right reserved.