首页 | 官方网站   微博 | 高级检索  
     

基于信噪比检验的病态总体最小二乘谱修正迭代法
引用本文:郭杰,郭淑妹,张宁.基于信噪比检验的病态总体最小二乘谱修正迭代法[J].西南师范大学学报(自然科学版),2020,45(3):10-15.
作者姓名:郭杰  郭淑妹  张宁
作者单位:信息工程大学 基础部, 郑州 450001
基金项目:国家自然科学基金项目(41174005,41474009).
摘    要:针对病态总体最小二乘谱修正迭代算法的不足,基于复共线性综合诊断,合理判断参数估计是否受到复共线性危害及危害程度,然后制定精准的正则化策略,提出了基于信噪比检验的病态总体最小二乘谱修正迭代算法,根据信噪比检验结果确定谱修正矩阵,根据改造的L曲线法确定谱修正参数.最后通过模拟算例验证了该算法求解病态总体最小二乘问题的数值稳定性和有效性.

关 键 词:总体最小二乘  病态模型  谱修正迭代  复共线性  信噪比
收稿时间:2018/8/21 0:00:00

Iteration Method of Correcting Characteristic Value to Ill-Posed Total Least Squares Based on Signal-to-Noise Ratio
GUO Jie,GUO Shu-mei,ZHANG Ning.Iteration Method of Correcting Characteristic Value to Ill-Posed Total Least Squares Based on Signal-to-Noise Ratio[J].Journal of Southwest China Normal University(Natural Science),2020,45(3):10-15.
Authors:GUO Jie  GUO Shu-mei  ZHANG Ning
Affiliation:Institute of Science, Information Engineering University, Zhengzhou 450001, China
Abstract:In this paper, the iteration method of correcting characteristic value and its improved algorithm have been analyzed in the detail, combined with its feature and shortcomings. Then the iteration method by correcting characteristic values to ill-posed total least squares based on signal-to-noise ratio been proposed. At last, some simulated examples been investigated to verify the numerical stability and effectiveness of parameter estimation of the proposed method.
Keywords:total least squares  ill-posed model  iteration method of correcting characteristic value  multicollinearity  signal-to-noise ratio
本文献已被 CNKI 等数据库收录!
点击此处可从《西南师范大学学报(自然科学版)》浏览原始摘要信息
点击此处可从《西南师范大学学报(自然科学版)》下载全文
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-23

京公网安备 11010802026262号