基于决策树模型的地铁线网短时OD客流预测

为了准确获取地铁线网短时OD(起讫点)的客流分布，从而高效协调运输能力和客运需求，结合集成学习思想构建了基于决策树模型的多时间粒度下地铁线网短时OD客流预测模型。首先利用地铁自动售检票数据分析得到线网OD客流出行的时空分布特征，引入多种时空影响因素对全网数据进行训练以及预测，其次分析了地铁线网OD客流量预测精度与时间粒度之间的关系，最后以苏州市地铁为对象进行实例分析。结果表明：相对于其他模型，研究模型不仅可以有效降低预测误差和拟合客流峰值，而且运算时间也节约了数倍，提高了地铁线网短时OD客流预测的准确性和效率。因此，所设计的模型可为地铁运营与控制系统提供重要数据，有助于运营者进行限流措施、行车计划等的制定与调整。

Short-time OD passenger flow prediction of subway line network based on decision tree model

In order to accurately obtain the short-time OD (origin-destination) passenger flow distribution of the subway line network, so as to efficiently coordinate the transportation capacity and passenger demand, a multi-time granularity short-time OD passenger flow prediction model of subway line network based on the decision tree model was constructed combined with the ensemble learning idea. Firstly, the spatial and temporal distribution characteristics of OD passenger flow on the line network were obtained by using the automatic ticketing data analysis, and various spatial and temporal influencing factors were introduced to train and predict the whole network data. Secondly, the relationship between the prediction accuracy and temporal granularity of OD passenger flow on the metro line network was analyzed. Finally, Suzhou Metro was taken as an example to carry out the case analysis. The results show that, compared with other models, the model can not only effectively reduce the prediction error and fit the peak passenger flow, but also save several times of computing time, which improves the accuracy and efficiency of the short-time OD passenger flow prediction of the subway line network. The designed model can provide important data input for the metro operation and control system, and help operators to formulate traffic restriction measures, travel plans and other strategies.