| Gaussian小波SVM及其混沌时间序列预测 |
| |
| 引用本文： | 郑永康,陈维荣,戴朝华,王维博.Gaussian小波SVM及其混沌时间序列预测[J].控制工程,2009,16(4). |
| |
| 作者姓名： | 郑永康 陈维荣 戴朝华 王维博 |
| |
| 作者单位： | 1. 西南交通大学,电气工程学院,四川,成都,610031
2. 西南交通大学,信息科学与技术学院,四川,成都,610031 |
| |
| 基金项目： | 西南交通大学博士研究生创新基金 |
| |
| 摘 要： | 为了提高混沌时间序列的预测精度,针对小波有利于信号细微特征提取的优点,结合小波技术和SVM的核函数方法,提出基于Gaussian小波SVM的混沌时间序列预测模型.证明了偶数阶Ganssian小波函数满足SVM平移不变核条件,并构建相应的Gaussian小波SVM.时混沌时间序列进行相空间重构,将重构相空间中的向量作为SVM的输入参量.用Ganssian小波SVM与常用的径向基SVM及Morlet小渡SVM进行对比实验,通过对Chen's混沌时间序列和负荷混沌时间序列的预测,结果表明,Ganssian小波SVM的效果比其他两种SVM更好.
|
| 关 键 词： | 混沌时间序列预测 相空间重构 Gaassian小渡核 负荷预测 |
Gaussian Wavelet SVM and Its Applications to Chaotic Time Series Forecasting |
| |
| ZHENG Yong-kang,CHEN Wei-rong,DAI Chao-hua,WANG Wei-bo.Gaussian Wavelet SVM and Its Applications to Chaotic Time Series Forecasting[J].Control Engineering of China,2009,16(4). |
| |
| Authors: | ZHENG Yong-kang CHEN Wei-rong DAI Chao-hua WANG Wei-bo |
| |
| Affiliation: | 1.School of Electrical Engineering;Southwest Jiaotong University;Chengdu;610031;China;2.School of Information Science and Technology;Sonuthwest Jiaotong University;Chengdu 610031;China |
| |
| Abstract: | To improve the accuracy of chaotic time series forecasting,Gaussian wavelet support vector machine(SVM)forecasting model is proposed,which combines the wavelet technology with SVM kernel function method,and based on that the wavelet is beneficial to extracting imperceptible features of signal. It is proved that the even order derivative Gaussian wavelet function is an admissible translation-invariant kernel of SVM,and corresponding Gaussian wavelet SVM is constructed. The chaotic time series is reconstructe... |
| |
| Keywords: | Chaotic time series forecasting phase space reconstruction Gaussian wavelet kernel load forecasting |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
