Improved Car-Following Model Considering Driving Styles and Changes in the Stimulus Term of the General Motors Model

The General Motors (GM) car-following model reflects the influence of the lead vehicle’s speed on the following vehicle, and it can effectively represent actual driving behavior. However, existing research indicates that using speed alone to represent the stimulus is unreasonable. In this study, the stimulus term of the GM model is modified by incorporating the difference between the gap between the lead and following vehicles and the desired gap for the following vehicle. During the study, the vehicle data from the next-generation simulation dataset are used as the research object. Through data processing, it was discovered that vehicles exhibit different driving styles, and this characteristic is incorporated into the car-following model. This leads to the development of the improved General Motors (IGM) car-following model. Through simulation experiments, this study confirms that the IGM model effectively captures the differences in driving styles. Finally, local stability, string stability analysis, and car-following experiments of the IGM model are conducted. Comparisons with the intelligent driver model car-following model demonstrate that the IGM model provides better car-following performance.