首页|基于FP-growth的老年行人交通事故损伤致因研究

基于FP-growth的老年行人交通事故损伤致因研究

扫码查看
原文链接
  • NETL
  • NSTL
  • 万方数据
为研究老年行人交通事故相关因素及其对损伤程度的影响,以中国交通事故深度调查数据库中2013-2023年710起机动车-老年行人事故为分析对象,利用FP-growth算法挖掘出4类影响因素28个类型变量共5 594项关联规则.研究结果表明:行人年龄、碰撞速度、事故时间段及事故地点是影响老年行人事故死亡率的显著因素.特别是,行人年龄和碰撞速度对死亡率有显著影响,随着碰撞速度的增加,死亡率显著上升;60岁及以上老年行人年龄每增加1岁,其死亡率提升0.037倍.此外,事故发生在夜间、郊区或村庄的死亡率更高.研究结果可为提高老年行人交通安全,制定相关安全措施提供一定参考.
Research on traffic accident injury causation of elderly pedestrian based on FP-growth
In order to study the factors related to the traffic accidents of elderly pedestrian and their influence on injury de-gree,710 vehicle-elderly pedestrian accident cases from the China In-Depth Accident Study(CIDAS)database from 2013 to 2023 were taken as the analysis object,and a total of 5594 association rules of 28 types of variables in 4 categories of influ-encing factors were excavated by the FP-growth algorithm.The results show that the pedestrian age,collision speed,accident time,and accident location are the significant factors affecting the death rate of elderly pedestrian accidents.In particular,the pedestrian age and collision speed have significant influence on the death rate.With the increase of collision speed,the death rate increases significantly.The death rate of elderly pedestrian over 60 years old increases 0.037 times for each additional year of age.In addition,the death rate is higher when the accident occurs at night,in the suburbs or in villages.The research results can provide certain reference for improving the traffic safety of elderly pedestrians and formulating the relevant safety measures.

FP-growth algorithmlogistic regression modelelderly pedestrian safetytraffic accident data analysis

刘永涛、张慧臣、袁诗泉、高隆鑫、王鹏、赵晨

展开 >

长安大学汽车学院,陕西西安 710064

中国汽车技术研究中心有限公司,天津 300300

中国安全生产科学研究院,北京 100012

FP-growth算法 逻辑回归模型 老年行人安全 交通事故数据分析

国家重点研发计划项目陕西省"两链"融合重点专项揭榜挂帅项目陕西省自然科学基础研究计划项目长安大学中央高校基本科研业务费专项资金项目

2021YFB25012022023JBGS-132023-JC-QN-0664300102223204

2024

10.11731/j.issn.1673-193x.2024.06.024

中国安全生产科学技术
中国安全生产科学研究院

中国安全生产科学技术

CSTPCD北大核心
影响因子：1.119
ISSN：1673-193X
年,卷(期)：2024.20(6)