车辆跟驰模型的发展综述  

Overview of the development of car-following models

作　　者：王道意 宇仁德[1] 闫兴奎 胡婧晖 朱燕华 崔淑艳 WANG Daoyi;YU Rende;YAN Xingkui;HU Jinghui;ZHU Yanhua;CUI Shuyan(School of Transportation and Vehicle Engineering,Shandong University of Technology,Zibo 255049,China)

机构地区：[1]山东理工大学交通与车辆工程学院,山东淄博255049

出　　处：《山东理工大学学报（自然科学版）》2022年第5期75-80,共6页Journal of Shandong University of Technology:Natural Science Edition

摘　　要：对车辆跟驰理论进行概述,分析车辆在非自由状态行驶下的跟驰特性。跟驰模型在近70年的发展历程中,吸引了交通工程学、统计物理学、心理学等不同领域的专家参与研究,建模时考虑因素由单一到多元、由简单到复杂,随着模型的不断发展,逐渐贴合于实际交通状况;分析了刺激-反应类模型、安全距离Gipps模型、优化速度FVD模型、元胞自动机模型,重点剖析了四类模型的建模思路和改进过程,对经典跟驰模型进行总结对比;展望了跟驰模型未来发展的趋势,指出在无人驾驶和智能网联的大环境下,基于机器学习的安全跟驰状态辨识方法研究、基于机器学习的跟驰特性研究及基于多元信息融合的跟驰模型构建等将是未来的发展方向。The car-following theory is summarized,and the car-following characteristics of vehicles in non free driving state are analyzed.During the nearly 70 year development process,the research of car-following model has attracted a large number of experts from different fields such as traffic engineering,statistical physics and psychology.The factors considered in modeling grow from single to multiple,from simple to complex.With the continuous development of the model,it gradually fits the actual traffic conditions.The review analyzes the stimulus-response model,the safety distance model,the optimized speed model,and the cellular automata model,focusing on the analysis of the modeling ideas and improvement processes of the four types of models.The classic car-following model is summarized and compared.In the end the future development trend of the car-following model is prospected.In the environment of unmanned driving and intelligent network connection,the research on safe car-following state identification method based on machine learning,the research on car-following characteristics based on machine learning and the construction of car-following model based on multiple information fusion will be the future development direction.

关 键 词：跟驰模型 交通工程 跟驰 建模 

分 类 号：U491[交通运输工程—交通运输规划与管理]