有序样本聚类方法在城市轨道交通运营时段划分中的应用

为合理划分轨道交通运营时段并指导其开行方案,提出一种基于有序样本聚类技术的运营时段划分方法。根据统计时段内客流数据,引入单向OD(origin-destination)概率矩阵,并给出单向OD概率矩阵的时序模型和提取方法;利用有序样本聚类方法,以最优分割法量化站间客流转移规律,求解聚类方案。最后以某一轨道交通线路为例,提取时间间隔为20 min的上行OD概率矩阵时间序列,以最优分割法进行聚类,将站间客流转移规律相近的统计时段归为一类,提出目标线路运营时段划分方案。

Application of Ordinal Clustering in the Division of Operation Periods for Urban Rail Transit

To divide rail transit operation period reasonably and conduct the operation plan,a method of operation period division using ordinal clustering was put forward in this paper.On the basis of one-way OD (Origin-Destination) probability matrix,the timing model and extraction method for one-way OD probability matrix in a statistical period was given.Then,the optimal partition algorithm was used to quantify interstation-passenger-transfer rules and solve the ordinal clustering scheme.Finally,the time sequences of uplink OD probability matrix (for an interval of 20 minutes) was constructed on the case of a rail transit line.According to the results of the ordinal clustering by the optimal partition algorithm,the statistical periods with similar interstation-passenger-transfer rules were classified together and the operation period division was proposed,providing decisionmaking basis for division of operation periods by the operation department.