现代交通技术2025，Vol.22Issue(2)：57-62,6.
基于自适应阈值的交通事故与交通违法行为时空关联分析
Adaptive Threshold-based Spatiotemporal Association Analysis of Traffic Accidents and Violations
摘要
Abstract
The research draws upon the optimization theory to construct a spatio-temporal association rule mining algorithm based on FP-growth,which incorporates adaptive threshold selection.The proposed method enables the flexible determina-tion of spatio-temporal thresholds between different associated objects,thereby producing more comprehensive mining results and avoiding the omission of certain covert but critical association relationships.Additionally,a more objective spa-tio-temporal threshold determination method can enhance the accuracy and reliability of association rule mining.A validi-ty analysis based on traffic violation data collected in Shenzhen from 2010 to 2020 demonstrates that the proposed method ex-hibits favorable stability and applicability.It can assist traffic managers in identifying the types of traffic violations that dem-onstrate a strong spatio-temporal association with accidents of varying severity,thereby providing references for the formula-tion of targeted traffic management measures.关键词
城市交通/时空关联规则/遗传算法/交通事故/交通违法Key words
urban traffic/spatio-temporal association rules/genetic algorithm/traffic accident/traffic violation分类
交通工程引用本文复制引用
王思,帅斌,张锐,黄文成..基于自适应阈值的交通事故与交通违法行为时空关联分析[J].现代交通技术,2025,22(2):57-62,6.基金项目
国家自然科学基金(72001179、72171198) （72001179、72171198）
