上海大学学报(自然科学版) ›› 2012, Vol. 18 ›› Issue (2): 174-177.doi: 10.3969/j.issn.1007-2861.2012.02.012

• 论文 • 上一篇    下一篇

基于内积矩阵Euv的最小二乘形式矩阵Padé-型逼近

潘宝珍,张明永,苏瑞   

  1. (上海大学 理学院,上海 200444)
  • 出版日期:2012-04-30 发布日期:2012-04-30

Least-Squares Matrix-Valued Padé-Type Approximation Based on Inner Matrix Euv

  1. (College of Sciences, Shanghai University, Shanghai 200444, China)
  • Online:2012-04-30 Published:2012-04-30

摘要: 为提高求矩阵Padé-型逼近解的精确度,给出一种求解矩阵Padé型逼近解的改进算法,即基于矩阵Euv的正交多项式Padé-型逼近算法.另外,当矩阵值幂级数展开式的系数产生微小摄动时,矩阵幂级数的Padé型逼近解变化往往很大,借助误差公式、内积单位矩阵和最小二乘法构造一种稳定性和精确度均有所提高的最小二乘形式矩阵Padé-型逼近算法.最后,对这两种算法分别给出完整的分子和分母行列式表达式.

关键词: Padé, 矩阵值, 线性泛函, 最小二乘法, 型逼近

Abstract: To obtain better results of matrix Padétype approximation, an improved algorithm is given for the matrix Padétype approximation, which is called an orthogonal polynomial algorithm of matrix Padétype approximation based on the matrix 〖WTHX〗E〖WTBX〗uv〖WTBZ〗 in the inner space. Matrix Padétype approximation can be quite sensitive to perturbation to the coefficients of the power series. In order to fill this gap, a leastsquares algorithm of matrix Padétype approximation is constructed. The numerator and denominator of the two algorithms determinant expressions are given.

Key words: least square, linear function, matrix-value, Padé-type approximation

中图分类号: