摘要
针对一票制公共交通IC卡刷卡系统存在的滞后时间问题,提出基于特征子序列修正的时差匹配推算方法。依托GPS车辆到站信息序列和刷卡聚类序列,分析首末站和中途站站间运行时间分布的差异化特征,分别从两序列中提取大时间差和连续小时间差等特征子序列;利用两种子序列间一对一或一对多的对应关系分别进行初步搜索和二次修正,兼顾匹配过程的效率和精度。将推算流程应用于山西省大同市公共汽车系统,结果显示,抽样车辆的刷卡匹配率均达到99%以上,并且针对不同类型的公共汽车线路具有较强的适应性。
To better solve the problem of lag time of IC card charging records in flat fare system, this paper proposes an estimation method for inferring time difference of IC card based on revision of characteristic subsequences. According to the arrival time sequence of GPS records and clusters of IC card charging records, the paper firstly analyzes the variation of travel time among stops which is for extracting a serial of large time difference and a serial of small time difference respectively. And then, the paper utilizes the relations of one-to-one or one-to-multiple to implement preliminary searching and revising by constraints of estimation efficiency and accuracy. The results in Datong show that the successful match ratio reaches to above 99% for sampled buses.
出处
《城市交通》
北大核心
2017年第1期59-65,共7页
Urban Transport of China
关键词
公共汽车
GPS
IC卡
刷卡聚类
首末站
中途站
滞后时间
特征子序列
修正
bus
GPS
IC card
clustering sequences of IC card charging records
terminals
bus stops
lag time
characteristic subsequences
revision
