基于GPS数据的出租车通勤识别及时空特征分析

Taxi Commute Recognition and Temporal-spatial Characteristics Analysis Based on GPS Data

为研究城市居民出行活动中使用出租车完成的通勤行为及其时空分布特征,基于出租车GPS设备所生成车辆活动轨迹数据,利用地理信息数据库处理及GIS技术,对出租车出行起讫点信息进行甄别,在交通小区划分的基础上,构建了出租车载客出行行为OD矩阵,提出了基于出租车出行的通勤客流识别模型,并据此建立了通勤距离和时长的计算模型,用于通勤行为的时空特征分析。最后,以西安市为实证对象,对建立的模型进行实证分析。研究结果表明:各交通小区间出租车出行交通流波动系数在0.2时较为稳定;工作日的早高峰通勤交通流为20 000~27 000车次,晚高峰通勤交通流为15 000~27 000车次,其中,周一和周二相对稳定,周三至周五略有增加;工作日出租车早、晚高峰平均载客车次数基本一致,周二出租车通勤人数最多,周四相对较少;整月内所有工作日早高峰出行中依靠出租车的通勤出行量约占单日高峰交通流均值的27%;出租车通勤出行平均距离为3.7km,通勤距离主要分布在2.5~7km,通勤出行的平均时长为17min,通勤时间主要分布在10~20min;就业地和居住地的识别结果与实际土地利用性质所代表的地块属性基本一致,该方法可对城市居民使用出租车方式的通勤行为以及就业地和居住地进行有效识别,并为城市通勤服务交通系统优化及空间活动的异质性分析提供依据。

In order to investigate the commute behavior by dint of taxi and temporal-spatial distribution characteristics of urban residents＇ travel activities,based on vehicle trajectory data generated by Taxi GPS device,the pick-up and drop-off information of commuters＇ travel by taxi was distinguished by geographic information database processing and GIS technology.On the basis of Traffic Analysis Zone（TAZ）division,the OD matrix of travel by taxi was established and the identification model of commute amount based on trajectory data was proposed.The compute model of commute time and distance was established to analyze temporal-spatial characteristics of compute behavior.Additionally,taken Xi＇an as an object of the research,the empirical analysis of established models was carried out.The results show that the fluctuationcoefficient of taxi traffic flow in each traffic zone is stable at 0.2.Average computing traffic flow of morning rush hours on weekdays is about 20 000 to 27 000 vehicles per day,and there are 15 000 to 27 000 vehicles per day in evening rush hours.Traffic flow in rush hours on Monday and Tuesday are relatively stable,with a slight increase from Wednesday to Friday.Average traffic flow is basically the same in morning and evening rush hours.The number of commuters by taxis on Tuesday is large,which becomes relatively less on Thursday.The scale of commuting travel by taxi in morning rush hours on weekdays during the whole month is about 27% of mean value of daily traffic flow in rush hours.The average commute distance of taxi is 3.7km and the commute distance is mainly in the range of 2.5km to 7km.The average commute time is 17 min and the commute time is mainly distributed between 10 mins to and 20 mins.The recognition results of employment place and residence area are consistent with the actual conditions of land use.The method presented in this paper is available to identify commute behavior,workplace,residence,with providing the basis for urban traffic system optimization of commuter services and heterogeneity analysis of spatial activities.