基于小波测量预处理的最小二乘估计  

Least Square Estimation Based On Wavelet Transform

作　　者：车金锐[1] 梁刚[2] 陈敏茹[3] 

机构地区：[1]中央国家机关工作委员会,北京100006 [2]河南大学计算机与信息工程学院,河南开封475001 [3]河南大学数学与信息科学学院,河南开封475001

出　　处：《数学的实践与认识》2006年第4期98-101,共4页Mathematics in Practice and Theory

摘　　要：以离散小波变换为工具,利用其特有的低通滤波属性和有效的抑制测量噪声的能力,将其与传统的最小二乘估计相结合,建立了一种基于小波测量预处理的改进最小二乘估计新方法,获得了比传统最小二乘估计好得多的估计结果;计算机仿真验证了算法的有效性.In this article, discrete wavelet transform is used as a tool to process data. By combining wavelet analysis with the traditional least square estimation, the traditional least square estimation is improved by using the low-pass filtering property and the ability of effectively restraining the measurement noises of wavelets. The method highly improved the validity of the traditional least square estimation. Computer simulation illustrated the feasibility and effectiveness.

关 键 词：最小二乘估计 小波变换 阈值处理 边界处理 

分 类 号：O241.5[理学—计算数学]