基于改进Apriori算法的高速公路交通事故关联分析
Correlation Analysis of Highway Traffic Accidents Based on Improved Apriori Algorithm
邱文利 1杨海峰 1张少波 1邱宇 2赵姣3
作者信息
- 1. 河北雄安京德高速公路有限公司,河北 保定 071000
- 2. 河北交投智能交通技术有限责任公司,河北 石家庄 050000
- 3. 长安大学 运输工程学院,陕西 西安 710064
- 折叠
摘要
针对现有交通数据中事故影响因素间的关联特性,为高速公路运营及管理部门提供精准化、细粒度的决策支持信息,该文考虑驾驶人、环境、道路和车辆4个维度,建立带约束的改进Apriori算法,挖掘影响高速公路交通事故的关联规则.在传统Apriori算法的基础上,规则前项和后项约束的增加,可以提高关联规则的准确性和挖掘效率.结合3 178条高速公路交通事故数据的分析结果表明:改进Apriori算法通过精准挖掘潜在因素和事故等级间的关联程度,降低无效关联规则数量,关联规则准确性、挖掘效率均大大提升.驾驶人性别、年龄、照明强度、车辆类型均与高速公路事故严重程度有强关联性;路面湿滑会使交通事故升级为一般事故;黑夜下的照明状况,是使轻微事故升级为一般和严重事故的主要因素.
Abstract
In order to explore the association characteristics between influencing factors of accidents from existing traffic data and provide precise and fine-grained decision support information for highway operation and management departments,a constrained improved Apriori algorithm was established to mine the association rules affecting highway traffic accidents by considering four dimensions:driver,environment,road,and vehicle.On the basis of the traditional Apriori algorithm,the improvement of antecedent and consequent constraints of the rule may increase the accuracy and efficiency of mining association rules.Based on the analysis of 3 178 highway traffic accident data,it is shown that the improved Apriori algorithm reduces the number of invalid association rules by accurately mining the correlation between potential factors and accident levels and improves the accuracy of association rules and mining efficiency.There is a strong correlation between driver gender,age,lighting intensity,vehicle type and the severity of highway traffic accidents.Wet and slippery road surfaces can escalate traffic accidents into general accidents.The lighting conditions under the dark night are the main factors that escalate minor accidents into general and serious accidents.
关键词
交通安全/高速公路/关联规则/Apriori算法/交通事故Key words
traffic safety/highway/association rules/Apriori algorithm/traffic accidents引用本文复制引用
出版年
2024
国家科技期刊平台
NSTL
万方数据