基于有序Logistic回归的驾驶人危险辨识影响因素分析

针对驾驶人危险辨识影响因素,开展危险辨识试验数据与自我报告数据的采集,采用k-means聚类分析方法将危险辨识时间划分为3类,进而基于有序Logistic回归分析方法检验不同影响因素的显著性.试验分析结果表明,危险情境特征对驾驶人危险辨识时间有影响,交通情境越混乱,所需时间越长,且交通情境越危险,所需时间相对越短;同时,驾驶人特征及其经历对驾驶人危险辨识时间有影响,男性的危险辨识时间较女性更短;摩托、电动车驾驶经历越丰富,危险辨识能力相对越好;驾龄、驾驶里程与自行车驾驶经历对危险辨识时间并无显著影响.环境因素和人口社会学因素对驾驶人危险辨识能力的影响是多方面的.

Influencing Factors Analysis of Driver's Hazard Perception Based on Ordered Logistic Regression

Aiming at the influencing factors of driver"s hazard perception, hazard perception experiments and self-reported data collection were carried out first, k-means clustering analysis method was used to divide the hazard perception time into three categories, and ordered Logistic regression analysis method was used to test the significance of the different factors. The results show that the characteristics of the dangerous condition have an impact on the driver"s hazard perception, the more clutter the traffic condition is; the longer the time required, the more dangerous the traffic condition is, and the shorter the time required; Drivers" characteristics and experience have an effect on hazard perception time. The time of males is shorter than that of females; The richer the driving experience of motorcycles or electric bicycles, the better the hazard perception; Driving age, mileage and bicycle driving experience have no significant effect on the hazard perception time. The influence of environmental factors and demographic factors on driver"s hazard perception is multifaceted.