Car-Following Model Based on Optimal Distance and Its Characteristics Analysis

To reveal the mechanism of car-following behaviors more accurately, a car-following optimal distance model (ODM) was proposed based on analysis of the optimal distance function, under an assumption that drivers in a car-following process always tend to maintain an optimal spacing distance. Then, the ODM and Gipps' car-following model were calibrated and evaluated using the next generation simulation (NGSIM) data. The numerical simulation method was adopted to analyze the acceleration characteristic of the ODM and its ability in reproducing the macroscopic phenomena of traffic flow. The results show that the simulation accuracy of the ODM in acceleration, velocity and distance MAE (mean average error) is 0.36 m/s2, 0.99 m/s and 0.73 m higher than that of the Gipps' model; the ODM can reproduce the stable state and shock wave phenomenon of traffic flow, and always tends to choose an acceleration that can make the distance between successive vehicles equal or just fluctuate about the optimal distance value in the stable state.