In order to better study the car-following characteristics in complex situations, the car-following behavior is compared to the result of the interaction of molecules in a one-dimensional pipeline. The existing car-following model based on molecular dynamics uses the demand safety interval and the expected vehicle speed as the main factors to establish the molecular following model, ignoring the influence of the relative speed between vehicles, and does not meet the actual car-following conditions. Therefore, this paper incorporates the relative velocity of the vehicle into the model structure, establishes the molecular interaction potential and the wall potential function, and constructs an improved molecular dynamics car-following model. The parameters in the car-following process are collected by high-precision vehicle-mounted instruments and the model parameters are calibrated by genetic algorithm. Finally, the accuracy of the model under different following states is verified separately and compared with the previous molecular model. The results show that the improved molecular following model can more effectively predict the following behavior of the vehicle.