首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于滚动时间窗的最小二乘支持向量机回归估计方法及仿真
引用本文:阎威武,常俊林,邵惠鹤.基于滚动时间窗的最小二乘支持向量机回归估计方法及仿真[J].上海交通大学学报,2004,38(4):524-526,532.
作者姓名:阎威武  常俊林  邵惠鹤
作者单位:上海交通大学,自动化系,上海,200030
基金项目:国家高技术研究发展计划(863)资助项目(2001AA413130)
摘    要:提出了一种基于滚动时间窗的最小二乘支持向量机(LSSVM)回归估计方法.该方法构造了滚动时间窗,利用滚动时间窗内的数据优化建模.模型随着时间窗的滚动进行在线更新,并对滚动时间窗内的数据分配不同的权值以充分利用数据的信息.将基于滚动时间窗的LSSVM回归估计方法应用于软测量建模.进行轻柴油凝固点的预估.结果表明,该建模方法十分有效.

关 键 词:最小二乘支持向量机  软测量  滚动时间窗  建模
文章编号:1006-2467(2004)04-0524-03

Least Square SVM Regression Method Based on Sliding Time Window and Its Simulation
YAN Wei-wu,CHANG Jun-lin,SHAO Hui-he.Least Square SVM Regression Method Based on Sliding Time Window and Its Simulation[J].Journal of Shanghai Jiaotong University,2004,38(4):524-526,532.
Authors:YAN Wei-wu  CHANG Jun-lin  SHAO Hui-he
Abstract:This paper introduced a least square SVM (LS SVM) regression method based on sliding time window. In this method, a sliding time window is built and data in the sliding time window are employed to construct the dynamic model of a system. The model of the system is refreshed on-line with the rolling of the time window. Different weights are assigned to the data in sliding time window in order to richly exploit the data information. The method was employed to construct soft sensor model, which is applied to the estimation of the frozen point of light diesel oil. The effective result indicates that the proposed method is a powerful method for the modeling.
Keywords:least square support vector machine (LS SVM)  soft sensor  sliding time window  modeling  
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号