主办:陕西省汽车工程学会
ISSN 1671-7988  CN 61-1394/TH
创刊:1976年

汽车实用技术 ›› 2023, Vol. 48 ›› Issue (22): 195-198.DOI: 10.16638/j.cnki.1671-7988.2023.022.038

• 标准·法规·管理 • 上一篇    

基于融合数据的事故预测

马 冲,于士杰,濮泽昱   

  1. 长安大学 汽车学院
  • 出版日期:2023-11-30 发布日期:2023-11-30
  • 通讯作者: 马 冲
  • 作者简介:马冲(2000-),男,硕士研究生,研究方向为交通安全数据分析,E-mail:gunipei1404@163.com。

Accident Prediction Based on Fusion Data

MA Chong, YU Shijie, PU Zeyu   

  1. School of Automotive, Chang'an University
  • Online:2023-11-30 Published:2023-11-30
  • Contact: MA Chong

摘要: 鉴于驾驶员驾驶行为和道路交通安全事故的极强相关性,驾驶员事故预测和风险评估 有利于对驾驶员分级管理和预警从而减少事故发生,而且在汽车保险公司、客货运营运企业 等领域具有广泛用途。文章基于以事故数据为主体的融合数据的事故及事故严重程度的预测 研究展开综述,首先介绍事故后果预测流程,接着介绍了众多学者在事故后果预测领域使用 的树模型包括决策树、随机森林、XGBoost,然后基于某市包含事故数据的融合数据实现并 对比三种模型,最后对事故后果预测研究进行展望。

关键词: 事故预测;融合数据;预测模型;实例分析

Abstract: In view of the strong correlation between drivers' driving behavior and road traffic accidents, driver accident prediction and risk assessment are conducive to hierarchical management and early warning of drivers, thus reducing accidents, and are widely used in automobile insurance companies, passenger and freight transport operating enterprises and other fields. This paper summarizes the research on accident and accident severity prediction based on fused data with accident data as the main body. First, it introduces the process of accident consequence prediction, then introduces the tree models used by many scholars in the field of accident consequence prediction, including decision tree, random forest and XGBoost, and then realizes and compares the three models based on fused data containing accident data in a city. Finally, it looks forward to the research on accident consequence prediction.

Key words: Accident prediction; Fusion data; Prediction model; Case analysis