Car-Following Model of Connected Vehicles with an Early Warning System Based on LTE-V Vehicular Communication

Meng-Yu Wang,Jiang-Feng Wang,Lei Chen,Hai-Qin Tang,Qing Xu

DOI: https://doi.org/10.1061/9780784482933.026

2020-01-01

Abstract:There are few studies on connected vehicle (CV) car-following models considering vehicle-to-vehicle communication (V2V) in actual traffic environments. This paper establishes in-field tests of car-following behaviors in different conditions based on LTE-V communication, which can warn following cars early. A novel car-following CV model with early warning system is proposed, suitable for complex traffic environments. LTE-V car-following behaviors are studied using in-field test data. Based on prediction accuracy analysis and goodness-of-fit of GM, FVD, and improved FVD models, a model is proposed based on the improved FVD model, which takes into account maneuver delay effect during acceleration information transmission and attenuates influence intensity of leading vehicle acceleration on the following vehicle, using communication delay as the attenuation factor via the law of information attenuation. Prediction effects of the improved FVD model and the proposed model are analyzed in different environments and velocities. Results show the proposed model has better accuracy, especially in low-velocity driving conditions or simple traffic environments.