| 引用本文: | 刘强,陆化普,王庆云,张红亮.基于支持向量机的区域运量滚动预测模型[J].哈尔滨工业大学学报,2011,43(2):138.DOI:10.11918/j.issn.0367-6234.2011.02.028 |
| LIU Qiang,LU Hua-pu,WANG Qing-yun,ZHANG Hong-liang.Rolling forecasting model of regional transportaion volume based on support vector machine[J].Journal of Harbin Institute of Technology,2011,43(2):138.DOI:10.11918/j.issn.0367-6234.2011.02.028 |
|
| 摘要: |
| 为寻求反映区域交通需求特性机理的运量预测方法,针对一般区域运量数据小样本的问题及其诱发因素的随机性和不可控制性,在分析区域交通需求特性及现有运量预测方法缺陷的基础上,采用以统计学习理论为基础的专门研究小样本情况下机器学习规律的支持向量机,建立了区域运量预测支持向量机模型.该模型通过预测值与统计值不断交互,实现区域运量的滚动预测,避免了建立和求解非线性函数的过程.以京津冀区域客运量预测为例,验证建立模型的合理性.结果表明,基于支持向量机的区域运量滚动预测较传统的预测方法提高了预测精度. |
| 关键词: 区域交通需求特性 区域运量 支持向量机 滚动预测 |
| DOI：10.11918/j.issn.0367-6234.2011.02.028 |
| 分类号:U491 |
| 基金项目:国家发展和改革委员会交通运输司资助项目 |
|
| Rolling forecasting model of regional transportaion volume based on support vector machine |
|
LIU Qiang1, LU Hua-pu1, WANG Qing-yun2, ZHANG Hong-liang3
|
|
1.Institute of Transportation Engineering,Tsinghua University,100084 Beijing,China;2.Department of Transportation,National Development and Reform Commission,100824 Beijing,China;3.Traffic and Transportation Department,Beijing Jiaotong University,100044 Beijing,China)
|
| Abstract: |
| To get the volume forecasting method which can reflect the mechanism of regional transportation demand characteristics,and considering the limited data of regional transportation volume and the stochastic and uncontrollable of the inducing factors,a rolling forecasting model of regional transportation volume was developed based on the Support Vector Machine and the analysis of the characteristics of regional transportation demand and the defects of existing forecasting methods.Through continuous interaction between predictive value and statistical value to update the training samples,the model realized the rolling forecasts of the regional transportation volume,and avoided the process of establishment and calculation of the nonlinear function.Taking the transportation volume of Beijing-Tianjin-Hebei area for instance,the rationality of the model has been validated and the results indicate that the model enhances the forecasting accuracy of regional volume,and has better prospects. |
| Key words: characteristics of regional transportation demand regional transportation volume support vector machine rolling forecasting |
