基于移动窗的多核最小二乘支持向量机建模算法

删除或更新信息，请邮件至freekaoyan#163.com(#换成@)

本站小编 Free考研考试/2024-01-16

-->

李琦,杜晓东,张洪略,邢丽萍.基于移动窗的多核最小二乘支持向量机建模算法[J].,2017,57(5):511-516

基于移动窗的多核最小二乘支持向量机建模算法

Multiple kernel LSSVM modeling algorithm based on moving window

DOI：10.7511/dllgxb201705011

中文关键词:动态建模核函数最小二乘支持向量机(LSSVM)移动窗

英文关键词:dynamic modeling kernel function least squares support vector machines(LSSVM)moving window

基金项目:国家自然科学基金资助项目(61403058)；中央高校基本科研业务费专项资金资助项目(DUT14LAB15).

作者	单位
李琦,杜晓东,张洪略,邢丽萍

摘要点击次数:1108

全文下载次数:955

中文摘要:

针对时变工业过程建模中存在的模型泛化性和适应性较低的问题，利用移动窗技术，通过使用多个核函数，提出了一种基于移动窗的多核最小二乘支持向量机(LSSVM)建模算法．该算法在最小二乘支持向量机算法基础上，利用多核组合代替单核，增强了模型的泛化能力；利用移动窗技术，增加了模型对时变工业过程的动态辨识能力及模型的更新效率．仿真实验结果表明，该算法具有更好的泛化性能．

英文摘要:

In order to solve the problem of modeling in the time-varying industrial process which has lower modeling generalization and adaptation, a multiple kernel least squares support vector machines (LSSVM) modeling algorithm based on moving window technology is proposed by using multiple kernel functions. Multiple kernel groups are used to replace single kernel in this algorithm based on least squares support vector machines, which can improve model′s generalization performance. The algorithm increases the dynamic recognition ability for the time-varying industrial process and updated efficiency of model through moving window technology. The simulation result shows that the algorithm has better generalization performance.

查看全文查看/发表评论下载PDF阅读器

关闭

基于移动窗的多核最小二乘支持向量机建模算法

本站小编 Free考研考试/2024-01-16

相关话题/

领限时大额优惠券,享本站正版考研考试资料!