基于多源数据的公交OD估计及客流分配模型

Transit OD Estimation and Assignment Model Based on Multi-source Data

文中基于互联网地图数据构建公交网络重构模型,实现站点间最少广义费用换乘路径的搜索;融合车辆位置及乘车交易数据构建公交OD估计模型,识别车辆到站时间及乘客上下车站点;构建基于公交网络和栅格起讫点的容量限制多路径分配模型,实现公交走廊的精准识别和客流预测;利用2019年9—10月的扬州市中心城区公交数据验证了模型的有效性,其中OD和出行链识别率分别达到60%和11%,公交走廊客流分配误差不超过10%.结果表明:模型能较好的反映城市公交出行的需求及客流特征.

Based on Internet map data,a bus network reconstruction model was constructed to realize the search of the least generalized cost transfer path between stations.The OD estimation model of public transport was constructed by fusing vehicle position and bus transaction data,and the arrival time of vehicles and the boarding and alighting stations of passengers were identified.A capacity-limited multipath allocation model based on bus network and grid starting and ending points was constructed to realize accurate identification of bus corridors and passenger flow prediction.The validity of the model is verified by the bus data of Yangzhou city center from September to October,2019,in which the recognition rate of OD and travel chain reaches 60%and 11%respectively,and the error of passenger flow distribution in bus corridors is less than 10%.The results show that the model can better reflect the demand and passenger flow characteristics of urban public transportation.