摘要
首先从微观层面上利用谱系聚类法对路口路段的拥堵程度进行分类,在此基础上利用基于节点删除的综合测度法对路口的重要程度进行判别,确定了相应的容量.然后从宏观层面上提出了判断整个路网拥堵程度的状态参量——最大流,并构造了节点和边都有容量约束的多源多汇最大流算法——改进的割集矩阵算法,最后利用算例验证了方法的可行性和有效性.
The paper first classifies the congestion levels of intersections and links into 3 classes using a hierarchical clustering method,which can not only distinguish the importance of various intersections but can be helpful in planning their capacities.Then it formulates a state parameter,named maximum flow,which may contribute in estimating the congestion level of network as a whole.Finally,the modified cutting-set matrix algorithm,namely multiple-input-multiple-output maximum flow method,is designed and the feasibility is proved by experimental results.
出处
《管理科学学报》
CSSCI
北大核心
2010年第11期35-40,共6页
Journal of Management Sciences in China
基金
国家自然科学基金资助项目(70671073)
国家高技术研究发展计划资助项目(2006AA11Z210)
关键词
谱系聚类
综合测度法
最大流
改进的割集矩阵法
hierarchical clustering method
comprehensive measurement method
maximum flow
improved cutting-set matrix method.
