摘要
地铁客流量是城市地铁交通运营组织的重要依据,客流随机性较大,其影响因素较多,因此加大了客流预测的难度。为了更加准确地预测城市地铁交通中的客流量,及时对客流组织方案进行调整,设计了一种基于非线性支持向量回归机的地铁客流量预测方法。该方法通过分析已采集数据的影响因素,确定对客流量影响较大的支持向量,然后构建预测模型进行预测。该模型可以通过调整影响因素的强度来提高预测精度。最后,通过算例验证:该方法可以有效地改善预测误差,适用于短期和不确定环境的地铁客流预测。
Subway traffic is one of the main basic data for subway operation and organization. However, the prediction of subway traffic is difficult for its randomness and multi-influencing factors. In this paper, an improved method of data prediction based on Support Vector Machine （SVM） is proposed to obtain a more precise prediction for subway-stations traffic, which can be used for operation and organization of subway. This method can predict subway traffic by analyzing collected data and determining which support vectors have more impact on traffic, and then adjust the strength of influencing factors to improve prediction accuracy. Experimental results show that SVM can evidently decrease error and predict subway traffic in a short time and in uncertain environment.
出处
《都市快轨交通》
北大核心
2014年第3期35-38,共4页
Urban Rapid Rail Transit
基金
教育部人文社会科学研究项目(13XJC630017)
关键词
地铁
支持向量机
地铁客流量
预测
影响因素
Metro
Support Vector Machine （SVM）
subway traffic
prediction
influencing factors
作者简介
赵钰棠，男，硕士研究生，从事城市轨道交通运营管理研究。zhaoyutang345@163．com
杨信丰，男，副教授，硕士生导师
