交通诱导信息发布策略的空间关联规则  

Spatial association rules of traffic guidance information release strategy

作　　者：樊玲玲[1] 关伟[1] 郭田德[2] 郑长青 关积珍 

机构地区：[1]北京交通大学城市交通复杂系统理论与技术教育部重点实验室,北京100044 [2]中国科学院大学数学科学学院,北京100049 [3]北京市公安局公安交通管理局,北京100037 [4]北京四通智能交通系统集成有限公司,北京100081

出　　处：《系统工程理论与实践》2014年第10期2688-2694,共7页Systems Engineering-Theory & Practice

基　　金：国家高技术研究发展计划(863)项目(2011AA110303);中央高校基本科研业务费专项基金(2012JBM073)

摘　　要：采用北京市可变信息板(variabl emessage signs,VMS)系统近三年发布的交通诱导信息数据,研究了交通诱导信息发布策略的空间关联规则.首先基于系统聚类算法分析事件点与上游VMS的关联度,根据其强弱将上游路段的影响空间分级,并确定不同影响等级空间范围的大小;然后采用基于关系代数理论的关联规则挖掘算法,计算位置相邻的VMS集合的支持度和置信度,确定不同影响等级空间范围内、相邻影响等级空间范围VMS集合的依存关系.实证分析表明:空间关联规则能将事件点上游的影响空间分级,并确定VMS间的依存关系,从而锁定诱导信息发布的各级影响空间以及相应的VMS,具有实际的应用价值,将应用于北京市交通诱导信息发布专家辅助决策系统.Based on traffic guidance information data released by Beijing variable message signs （VMS） system in recent three years, spatial association rules of traffic guidance information release strategy are studied in this paper. At first, hierarchical clustering algorithm is adopted to analyze association degree between event node and upstream VMS. According to intensity of association degree, upstream road sections of event node are divided into different levels of influence space. At the same time, spatial extent of each influence space is determined. Then, association rule mining algorithm based on relational algebra theory is adopted to analyze association degree between adjacent VMS sets. Empirical analysis is conducted based on traffic guidance information data released by Beijing VMS system. Empirical analysis shows that spatial association rules can classify upstream influence road sections into different levels, and identify dependencies between VMS sets. The spatial association rules will be used to lock different influence road sections and corresponding VMS sets in traffic guidance information dissemination. The proposed spatial association rules are applicable to develop traffic guidance information release strategy.

关 键 词：交通诱导信息 发布策略 影响空间分级 依存关系 

分 类 号：U491[交通运输工程—交通运输规划与管理]