Over the past decades, the intense emphasis has been placed on the understanding of car-following behavior and the factors that affect it. The car-following process, however, still remains a very complex field of study in spite of all the efforts. This paper focuses on the study of the effect that the class of the vehicle, leading heavy vehicles in particular, causes on the following vehicle behavior, specifically in terms of the following bumper-to-bumper distance (gap) that the following vehicle keeps from the lead vehicle. This was done by extracting and analyzing different car following episodes in the Next Generation Simulation (NGSIM) dataset for Interstate 80 (I 80) in Emeryville, California, USA. The results of the statistical analysis are compared to that of the synthesized literature of research efforts that have been carried out on the topic, then are further assessed utilizing different calibrated behavioral clusters for the Gazis-Herman-Rothery (GHR) car-following model to address the similarities and differences in car-following behavior between drivers of the same vehicle class. The paper also validates the results of the statistical analysis and highlights possible future implementations.

中文翻译：

---

基于车辆类别、速度和道路几何形状的驾驶员行为识别和分类

在过去的几十年里，人们高度重视对跟车行为及其影响因素的理解。然而，尽管做出了所有努力，跟车过程仍然是一个非常复杂的研究领域。本文重点研究车辆的类别，尤其是重型车辆，对跟随车辆行为的影响，特别是跟随车辆与保险杠之间的距离（间隙）。领头的车辆。这是通过在美国加利福尼亚州埃默里维尔的 80 号州际公路 (I 80) 的下一代模拟 (NGSIM) 数据集中提取和分析不同的汽车跟随事件来完成的。将统计分析的结果与针对该主题进行的研究工作的综合文献进行比较，然后使用不同的校准行为集群进一步评估 Gazis-Herman-Rothery (GHR) 跟驰模型解决同一车辆类别的驾驶员在跟车行为方面的异同。该论文还验证了统计分析的结果，并强调了未来可能的实施。