计算机工程与应用 ›› 2013, Vol. 49 ›› Issue (16): 240-243.

• 工程与应用 • 上一篇    下一篇

冥函数变换在短时交通流组合预测中的应用

李  宁1,2,王晓东1,2,侯俊峰3,黄国勇1,2   

  1. 1.昆明理工大学 信息工程与自动化学院,昆明 650500
    2.云南省矿物管道输送工程技术研究中心,昆明 650500
    3.中烟工业有限责任公司,河南 许昌 461000
  • 出版日期:2013-08-15 发布日期:2013-08-15

Application of power function on short-term traffic flow prediction

LI Ning1,2, WANG Xiaodong1,2, HOU Junfeng3, HUANG Guoyong1,2   

  1. 1.Faculty of Information Engineering and Automation, Kunming University of Science and Technology, Kunming 650500, China
    2.Engineering Research Center for Mineral Pipeline Transportation, Kunming 650500, China
    3.China Tobacco Henan Industrial CO, LTD, Xuchang, Henan 461000, China
  • Online:2013-08-15 Published:2013-08-15

摘要: 实际交通流是一个明显含有噪声的非线性时间序列。针对这一特点提出对此时间序列进行冥函数变换,变换之后的噪声会比原始信号的压缩程度更大,从而降低白噪声对预测结果的不利影响;利用最小二乘支持向量机(LS-SVM)对自回归求和滑动平均(ARIMA)模型的预测结果进行循环补偿;通过冥函数反变换对输出结果进行相应的信号还原。实验预测结果表明,经过冥函数变换后的组合预测模型具有较高的预测精度。

关键词: 冥函数变换, 自回归求和滑动平均模型(ARIMA), 最小二乘支持向量机(LS-SVM), 短时交通流预测

Abstract: Traffic flow is a non-linear time series with obvious noise. For this feature, a conversion method using power function is proposed to the time series, after the transformation, the noise?than the original signal?will be?a greater degree of?compression. And reduce the adverse effects of the predicted results. Use LS-SVM to compensation loop for the prediction?of? ARIMA. Use the inverse transform method to restore the signal?output. The results show that the combination model based on the conversion of power function possesses satisfactory accuracy.

Key words: power function, Autoregressive Integrated Moving Average Model(ARIMA), Least Squares Support Vector Machines(LS-SVM), short-term traffic flow prediction