基于分层Logit模型的车辆换道行为研究  

Study on Vehicles Lane-changing Behavior Based on Nested Logit Model

作　　者：曲大义[1] 李娟[1] 刘聪[1] 万孟飞 曹俊业[1] 

机构地区：[1]青岛理工大学汽车与交通学院,青岛266520

出　　处：《科学技术与工程》2017年第5期307-311,共5页Science Technology and Engineering

基　　金：国家自然科学基金项目(51178231)资助

摘　　要：研究分析复杂交通场景中车辆换道行为,揭示车辆运行特性及其规律。基于效用理论方法,以实现驾驶行为决策效用最大化为目标,建立车辆换道分层Logit模型。系统分析影响车辆行为变化的因素,建立各层次参数变量相关模型。仿真分析表明:较一般的车辆行为模型,车辆换道分层Logit模型更能准确地描述实际交通场景中车辆换道的行为决策过程,有效提高车辆运行效率。研究结果为智能车路协同与交互行为、车辆可变限速技术、自适应巡航控制技术等提供理论依据和技术支撑。Lane-changing behavior of vehicle in complex traffic scenes is analyzed to reveal the characteristics and principles of vehicle operation. The Logit model of vehicle lane-changing behavior is developed based on the utility theory aiming at achieving the maximum decision-making utility. The factors influencing behavioral changes of vehicles are systematically analyzed and relevant models of parameter variables for each layer are developed. Simulation analysis shows that compared with general behavioral models of vehicles, the nested Logit model of vehi-cle lane-change can more accurately describe decision-making process of lane-changing behavior in the actual traffic scenes and greatly improve efficiency of vehicle operation. The research results will provide theoretical basis and technical support for intelligent vehicle infrastructure cooperation and interaction behavior, variable speed-limit technology of vehicles and adaptive cruise control technology.

关 键 词：交通运输系统工程 换道行为 效用理论 分层Logit模型 

分 类 号：U491.2[交通运输工程—交通运输规划与管理]