多源数据融合驱动的居民出行特征分析方法

结合传统抽样调查数据和交通大数据，研究多源数据融合驱动的居民出行特征分析方法.根据传统入户抽样调查居民的年龄结构、职业、车辆拥有、人口，以及手机信令数据分析出行频次分布等因素进行综合分析，获取居民初步出行特征；基于手机信令、IC、AFC、GPS 等大数据，通过出行时间分布、OD分布和出行方式结构对居民的出行特征进行综合矫正分析；最后，以广州市为例进行实证分析.对比研究传统抽样调查和多元数据融合分析方法可知，传统抽样调查居民出行漏报率为30%，每天出行2次的比例相差39.5%，全方式非通勤出行比例、晚高峰公交和地铁出行比例分别相差7.4%、8.1%和12.6%.结果表明，多源数据融合驱动的居民出行特征分析方法，在总量上有效挖掘居民出行的沉默需求，在时空分布上起到了“削峰填谷”的作用，是一种研究居民出行特征的有效方法.

Resident Travel Characteristics Analysis Method Based on Multi-source Data Fusion

This study proposes an analysis method to investigate residents' travel characteristics based on multisource data fusion by using traditional household travel survey and transportation big data. The residents travel characteristics were initially analyzed through a combination of traditional household survey data analysis (age, occupation, vehicle ownership, population distribution) and mobile phone signaling data analysis (travel frequency distributions). Then, the residents' travel characteristics were further analyzed through mobile phone signaling, Intelligent Card (IC), Automatic Fare Collection (AFC), the Global Positioning System(GPS) and other big data. The analysis results include residents' travel time distribution, Origin- Designation (OD) distribution, and travel mode structure. The resident travel characteristics of Guangzhou, China was analyzed as an example. The study then compared the proposed method with traditional household survey data analysis methods. The results indicated that about 30% residents' trips were not recognized by traditional household sampling surveys; the proportion of travel rate of twice a day generated by these two methods were 39.5% different; and the differences generated by these two methods for non-commuting trips, bus trips in PM peak hours, and subway trips in PM peak hours were respectively 7.4%,8.1%and 12.6%. Compared with the traditional method, the multi- source data fusion analysis method is more effective to identify and analyze residents travel characteristics. It plays an important role to examine and balance residents' travel needs with the time and space distributions.