Car-following modeling based on Morse model with consideration of road slope in connected vehicles environment

To develop connected and automated transportation system, it is essential to model the car-following behavior of connected vehicles (CVs). In recent years, car-following models based on potential fields have attracted increasing attention owing to their objectivity, universality, variability, and measurability. However, existing potential-field-based car-following models do not consider unified gravity and repulsion between vehicles, and they lack scalability to other driving behaviors, including lane-changing and overtaking. In this study, we proposed a multiple risk potential-field-based car-following model (MRPFM) for traffic flow in CVs environment, which integrated multiple risk potential fields of traffic subjects, including road markings, road longitudinal slopes, and vehicle interactions. This model revealed the relationships between the potential field, interaction force, and driving risk. In particular, the Morse model was applied to formulate the risk potential field of vehicle interaction, and the risk potential field of the road longitudinal slope was derived using force analysis. The experiments were conducted based on the Zen Traffic Data dataset, which contained precise and complete trajectories of all vehicles along a 2 km freeway segment of longitudinal slope. We conducted a comparative analysis between the MRPFM and four prevailing car-following models-that is, the optimal velocity model, the full velocity difference model, the intelligent driver model, and the driving risk potential-field model. The results showed that the MRPFM achieved the best performance in terms of accuracy and stability.& COPY; 2023 Elsevier B.V. All rights reserved.