驾驶风格和分心状态对车道偏离预警系统效用的影响  

The Impact of Driving Style and Distracted State on the Utility of Lane Departure Warning System

作　　者：张名芳 王子茜 李坤[1] 王力[1] 王庞伟[1] ZHANG Mingfang;WANG Zixi;LI Kun;WANG Li;WANG Pangwei(Beijing Key Laboratory of Urban Road Intelligent Traffic Control Technology,North China University of Technology,Beijing 100144,China)

机构地区：[1]北方工业大学城市道路智能交通控制技术北京市重点实验室,北京100144

出　　处：《交通工程》2024年第5期1-8,30,共9页Journal of Transportation Engineering

基　　金：国家重点研发计划项目(2022YFB4300400);北京市教育委员会科学研究计划项目(KM202210009013);中乌合作专项(106051360024XN017)。

摘　　要：为探究可匹配不同驾驶风格和分心状态水平的车道偏离预警模式,开发可满足不同用户驾乘体验需求的车道偏离预警系统(LDW),通过K-means聚类的高斯混合模型将被试分为保守型、普通型和激进型,设计不同难度等级分心次任务,针对早、晚期2种预警提前期模式开展驾驶模拟试验。基于驾驶数据,采用K-M非参数生存分析方法生成单变量影响下偏离持续时间的生存函数,在此基础上建立偏离持续时间COX比例风险回归模型进行多变量影响分析,并设计基于技术接受度模型的主观评价方法。结果表明:驾驶风格和分心状态对车道偏离持续时间产生显著影响;驾驶风格越激进,分心等级越高,偏离持续时间越长;同时,不同驾驶风格和分心状态对应的驾驶偏离风险和主观接受度均存在差异,综合主客观结果得出保守型、普通型和激进型驾驶人的最佳预警策略,研究成果可为LDW系统个性化设计提供基础支撑。To explore departure warning models that can be applied to different driving styles and distracted states and further develop Lane Departure Warning(LDW)systems to meet the various driving experience need of the drivers,the Gaussian mixture model with K-means clustering was used to classify as“conservative”,“normal”and“aggressive”and different difficulty level distraction sub-tasks were designed to carry out driving simulation experiments for the two warning advance modes,early and late.Based on driving data,K-M non-parametric survival analysis method was used to generate the survival function of the departure duration under the influence of various key variables.Furthermore,the departure duration COX proportional risk regression model was developed for multivariate impact analysis.the COX proportional risk regression model of departure was constructed.By establishing technology acceptance modeling,subjective evaluation method was further constructed.The results show that driving style and distraction state have a significant effect on lane departure duration.The more aggressive the driving style and the higher the distraction level,which will result in a longer duration of departure.At the same time,there are differences in the risk and subjective acceptance of driving departure under different driving styles and distraction states.The optimal warning strategies for conservative,normal and aggressive drivers are derived by combining the subjective and objective results.The result can provide the basic support for the personalized design of the LDW system.

关 键 词：智能交通 车道偏离预警系统 驾驶风格 分心状态 偏离持续时间 生存分析 

分 类 号：U467.1[机械工程—车辆工程]