中文摘要：

公交客流规模测算往往存在调查成本受限和准确度要求较高的矛盾。提出基于公交IC卡历史数据与人工补充调查数据的数据融合测算方法,以准确推算公交客流规模。首先根据公交线路的基本属性,采用聚类分析方法划分线路类型,从每一类中选择具有代表性的线路。基于IC卡数据分析公交客流时变特征,运用有序样本聚类Fisher算法将线路小时刷卡量进行聚类分析。划分刷卡量相似时段,进而采用优化方法确定调查抽样率,确定相应的调查车辆进行人工补充调查,最终经过数据融合计算获得公交客流规模。基于上海市某辖区IC卡数据进行案例分析,测算得到三类公交线路的日均客流量。

英文摘要：

Accurate passenger flow estimation through surveys does not come without costs. This paper proposes a data fusion method based on the data from public transit IC card and supplementary surveys to accurately estimate passenger flow information. This paper first divides bus service routes into groups by their characteristics using cluster analysis method, and then selects one representative route from each group. Based on the temporary variation of bus passenger flow extracted from IC cards data, the paper categorizes IC card charging records per hour using Fisher algorithm of ordered sample cluster. By grouping time periods with similar IC card charging volumes, the paper determines the optimized sample rate and corresponding buses for the supplementary surveys. Consequently, bus passenger flows are estimated by data fusion method. Taking one district in Shanghai as an example, the paper demonstrates how to estimates daily passenger volumes of three types of bus routes using the above method.