摘要
城市交通流具有复杂性、时变性和随机性,实时准确的交通流量预测是实现智能交通诱导及控制的前提.综合分析交通流量影响因素的基础上,进行多路段的交通流量预测研究,提出了基于最小二乘支持向量机的交通流量预测改进模型,并应用平安大街的流量数据进行实例验证.结果表明,该模型具有学习速度快、跟踪性能好及泛化能力强等优点,在交通流预测中更具有实用性和推广性.
With the complexity,time variation and randomness of urban traffic flow,the issue of real-time and accurate traffic volume forecasting is very essential to the intelligent traffic guidance,control,and management.Including synthetic analysis of factors affecting the traffic volume and research on traffic volume forecasting,an improved traffic volume forecasting model based on LS-SVM is given,and a case study applying Ping-An Avenue traffic flow data is carried out to validate the model.The results indicate that this model features the high learning speed,good approximation and strong generalization ability,and thus it＇s more practical and easier to promote the traffic volume forecasting.
出处
《北京交通大学学报》
CAS
CSCD
北大核心
2011年第2期114-117,136,共5页
JOURNAL OF BEIJING JIAOTONG UNIVERSITY
基金
国家自然科学基金资助项目(60874079)
教育部重点资助项目(108127)
轨道交通控制与安全国家重点实验室自主课题项目资助(RCS2009ZT003)
关键词
最小二乘支持向量机
交通流量
实时预测
多路段
least squares support vector machine
traffic volume
real-time forecasting
multi sections
