多元数据下的公交站点客流不确定性分析

利用长期收集的公交站点IC卡刷卡数据，基于区间不确定性理论，提出公交上下车站点区间不确定性客流推导方法.首先，将公交IC卡数据和GPS数据与公交站点信息相融合，确定IC卡刷卡的上车站点及其区间不确定性客流；然后，对公交刷卡行为进行分析，考虑乘客个体出行特征、乘客出行距离和站点吸引权重，提出下车站点客流推导概率模型及相应算法，通过区间数处理方法，得到下车站点客流区间值.最后，以2015年11月13日~12月25日深圳市21路公交为期6周的IC卡刷卡数据和GPS数据为例进行实例分析.

Uncertainty Analysis to Passenger Flow of Bus Stations Based on Multivariate Data Fusion

Using the data of smart card of bus station collected in the long term, a method of the interval uncertainty passenger flow derivation of the bus station is proposed based on the interval uncertainty theory. On condition that the bus smart card data and GPS data as well as bus station information are fused, boarding stations of bus passengers are identified, and then interval uncertainty passenger flow of boarding stations are inferred. By analyzing the behavior of bus swiping card, and considering the individual trip characteristics and the travel distance of passengers to attract the weight of the station, the probabilistic model of alighting passenger flow and its corresponding algorithm are proposed, and the interval value of the passenger flow at the alighting station is obtained according to interval number processing method. An example is given and analyzed by taking smart card data and GPS data of No. 21 bus in Shenzhen City from November 13, 2015 to December 25, 2015.