修正的高阶Hermite—Fejer插值的加权L^p逼近

本文给出了基于Ｃｈｅｂｙｓｈｅｖ结点的高阶Ｈｅｒｍｉｔｅ－Ｆｅｊｅｒ插值多项式的两种修正形式，并证明了这两种修正对ｆ∈Ｌｗ＾ｐ均可给出逼近阶ｗ（ｆ，１／ｎ）ｐ．同时文中也给出了基于Ｃｈｅｂｙｓｈｅｖ结点的Ｈｅｒ－ｍｉｔｅ－Ｆｅｊｅｒ及Ｈｅｒｍｉｔｅ插值多项式对Ｃ［－１，１］及Ｃ＾ｒ［－１，１］类函数的逼近阶。     

关于Hermite-Fejer插值过程的平均收敛

文［１］讨论了某些非Ｗ－过程的插值算子的加权平均逼近的收敛性和收敛阶．如记Ｈｎ（ｆ；ｘ）为以第二类Ｃｈｅｂｙｓｈｅｖ多项式Ｕｎ（ｘ）的零点作为插值节点，区间［－１，１］上的函数ｆ（ｘ）的Ｈｅｒｍｉｔｅ－Ｆｅｊｅｒ插值算子，［１］中证得：定理Ａ当０＜ｐ...     

有理插值的基本特征

１　引　言 记 ｎ为次数不超过ｎ的一元多项式函数类,约定零多项式的次数为－∞,即ｄｅｅ（０）＝－∞;记 ｍ,ｎ为分子属于 ｍ,分母属于 ｎ＼｛０｝的一元有理函数类．我们约定：本文所采用的概念和记号将与文［１］保持一致,本文中“有理插值问题”系指文［１］中“有理插值问题（２．１）（２．２）”,并简记为ＲＩＰ． 文［１］在［２－８］的基础上引进了ＲＩＰ的　方程组,定义了　插值式,指出了其与经典结果的关系,这为我们分析ＲＩＰ提供了一个有力的工具．本文将在文［１］的基础上,深入讨论ＲＩＰ的基本特征．在本节和下…     

数据插值和共轭函数

设Ｇ（ｚ）在｜ｚ｜＜ρ（ρ＞１）中解析，且数据Ｒｅ［Ｇ（ｅｊ２ｋπ／ｎ）］；ｋ＝０，１，…，ｎ－１已给出，其中ｎ＝２ν＋１，本文构造了一个ν次多项式Ｐν（ｚ）满足插值条件Ｒｅ［Ｐν（ｅｊ２ｋπ／ｎ）］＝Ｒｅ［Ｇ（ｅｊ２ｋπ／ｎ）］，ｋ＝０，１，…，ｎ－１．并估计了误差‖Ｇ（ｅｊω）－Ｐν（ｅｊω）‖．此外，还给出了一个Ｗａｌｓｈ类型的超收敛定理．     

Hermite型插值算子对可微函数的逼近

Ｈｅｒｍｉｔｅ型插值算子对可微函数的逼近章仁江（中国计量学院，杭州３１００３４）关键词Ｈｅｒｍｉｔｅ型插值算子，Ｊａｃｏｂｉ多项式．分类号ＡＭＳ（１９９１）４１Ａ／ＣＣＬＯ１７４设（１）＞ｘ１＞ｘ２＞…＞ｘｎ＞（－１），ｘｋ＝ｃｏｓθｋ（ｋ＝１，２，...     

P．Turán问题XXXV的一个注记

本文给出了有关Ｐ．Ｔｕｒáｎ，问题ＸＸＸＶ［关于逼近论的某些未解决的问题，Ｊ．ＡｐｐｒｏｘｉｍａｔｉｏｎＴｈｅｏｒｙ，１９８０，２９（１）：２３－８５］的一个结果．设ｒ＿（ｉｎ）（ｘ）为（０，２）插值的第一类基函数，其插值节点为（１－ｘ）（ｘ）之零点而Ｐ＿ｎ（ｘ）为ｎ次Ｌｅｇｅｎｄｒｅ多项式．那么．但对ｆ￣＊＝ｘ￣２却有     

关于一个矩阵迹的不等式的注记

本文修正了［２］中的一个矩阵迹的不等式的一些错误，证明了ｔｒ［（Ａａ一Ｂａ）（Ａ一β一Ｂβ）］＜０当且仅当αβ＞０且Ａ≠Ｂ，ｔｒ［（Ａａ－Ｂａ）（Ａ－β－Ｂ－β）］＞０当且仅当αβ＜０且Ａ≠Ｂ，这里Ａ，Ｂ是ｎ×ｎ的Ｈｅｒｍｉｔｅ正定矩阵．     

再论求导数零点的二次收敛迭代法

一维搜索是最优化理论数值计算的一个基本问题,它可归结为求定义在开凸区域Ｄ上的可微函数 ｆ的导数零点．若用 Ｎｅｗｔｏｎ法求导数零点,则涉及到二阶导数的计算．若用带导数的三次插值法则需要开平方的计算［１］．为了克服上述问题,本文作者之一在 １９７９年［２］首次提出了下述具有二阶收敛速度的迭代法：通常,我们称迭代法（０．１）为基于信息集（ｆ（ｘｎ）,ｆ’（ｘｎ）,ｆ（ｘｎ－１）,ｆ’（ｘｎ－１）｝的迭代法,而δ（ｆｘｙ）是基于信息集｛ｆ（ｘ）,ｆ＇（ｘ）,ｆ（ｙ）,Ｆ＇（ｙ））｝的三次插值多项式在ｘ处…     

P.Turan问题XXXV的一个注记

本文给出了有关Ｐ．Ｔｕｒａｎ问题ＸＸＸＶ［关于逼过论的某些未解决的问题，Ｊ．Ａｐｐｒｏｘｉｍａｔｉｏｎ Ｔｈｅｏｒｙ，１９８０，２９（１）：２３－８５］的一个结果。设ｒｉｎ（ｘ）为（０，２）插值的第一类基函数，其插值节点为（１－ｘ）Ｐｎ＇（ｘ）之零点而Ｐｎ（ｘ）为ｎ次Ｌｅｇｅｎｄｒｅ多项式。那么ｍａｘ－１≤ｘ≤１∑ｉ＝１ｎ│ｒｉｎ（ｘ）│＝Ｏ（ｎ＾５／２ｌｎｎ）．但对ｆ＾＊＝ｘ＾２却有ｌｉｍ↓ｎ→     

一类滞后生态微分方程的初值问题

对一般的滞后系统，人们采用了将滞后变量x(t-1)用一个Hermite插值多项式来处理，从而把滞后系统转化为常微分方程系统来求其数值解(见文[2]，[3])。本文根据[2]中的表1选用了一个带有五次Hermite插值多项式的四阶Runge-Hum法来求两个常见的滞后初值问题．     

Hermite interpolation with symmetric polynomials

We study the Hermite interpolation problem on the spaces of symmetric bivariate polynomials. We show that the multipoint Berzolari-Radon sets solve the problem. We also give a Newton formula for the interpolation polynomial and use it to prove a continuity property of the interpolation polynomial with respect to the interpolation points.     

About measures and nodal systems for which the Hermite interpolants uniformly converge to continuous functions on the circle and interval

We obtain the Laurent polynomial of Hermite interpolation on the unit circle for nodal systems more general than those formed by the n-roots of complex numbers with modulus one. Under suitable assumptions for the nodal system, that is, when it is constituted by the zeros of para-orthogonal polynomials with respect to appropriate measures or when it satisfies certain properties, we prove the convergence of the polynomial of Hermite-Fejér interpolation for continuous functions. Moreover, we also study the general Hermite interpolation problem on the unit circle and we obtain a sufficient condition on the interpolation conditions for the derivatives, in order to have uniform convergence for continuous functions.Finally, we obtain some improvements on the Hermite interpolation problems on the interval and for the Hermite trigonometric interpolation.     

Hermite插值多项式在不同基下的显式表示

本文利用对偶基的概念,导出了 Ｈｅｒｍ ｉｔｅ 插值多项式在不同基下的显式表示,这给人们对 Ｈｅｒｍ ｉｔ插值多项式在不同基下从一种表示转换到另一种表示带来极大的方便     

一类收敛的线性Hermite重心权有理插值

众所周知, Hermite有理插值比Hermite多项式插值具有更好的逼近性, 特别是对于插值点序列较大时, 但很难解决收敛性问题和控制实极点的出现. 本文建立了一类线性Hermite重心有理插值函数$r(x)$,并证明其具有以下优良性质: 第一, 在实数范围内无极点; 第二, 当$k=0,1,2$时,无论插值节点如何分布, 函数$r^{(k)}(x)$具有$O(h^{3d+3-k})$的收敛速度; 第三, 插值函数$r(x)$仅仅线性依赖于插值数据.     

Newton basis for multivariate Birkhoff interpolation

Multivariate Birkhoff interpolation is the most complex polynomial interpolation problem and people know little about it so far. In this paper, we introduce a special new type of multivariate Birkhoff interpolation and present a Newton paradigm for it. Using the algorithms proposed in this paper, we can construct a Hermite system for any interpolation problem of this type and then obtain a Newton basis for the problem w.r.t. the Hermite system.     

An Improvement to the Computing of Nonlinear Equation Solutions

In this paper, we suggest an improvement to the iterative methods based on the inverse interpolation polynomial, also referred to as the generalized Hermite interpolation, which increases the local order of convergence. A symbolic computation allows us to find the best coefficients with regard to the order of convergence. The adaptation of the strategy presented here gives a new iteration function with a new evaluation of the function. It also shows a smaller cost if we use adaptive multi-precision arithmetic. The numerical results computed using this system, with a floating point system representing 200 and 1000 decimal digits support this theory.     

On the hermite matrix of a polynomial

We investigate the location of the eigenvalues of the Hermite matrix of a given complex polynomial, the question under what conditions a given polynomial and the characteristic polynomial of its Hermite matrix are identical, and the question under what conditions the Hermite matrix has only one distinct eigenvalue.     

Convergence rates of a family of barycentric osculatory rational interpolation

It is well-known that osculatory rational interpolation sometimes gives better approximation than Hermite interpolation, especially for large sequences of points. However, it is difficult to solve the problem of convergence and control the occurrence of poles. In this paper, we propose and study a family of barycentric osculatory rational interpolation function, the proposed function and its derivative function both have no real poles and arbitrarily high approximation orders on any real interval.     

Polynomials associated with a functional product of the Hermite type

We have found the motivation for this paper in the research of a quantized closed Friedmann cosmological model. There, the second‐order linear ordinary differential equation emerges as a wave equation for the physical state functions. Studying the polynomial solutions of this equation, we define a new functional product in the space of real polynomials. This product includes the indexed weight functions which depend on the degrees of participating polynomials. Although it does not have all of the properties of an inner product, a unique sequence of polynomials can be associated with it by an additional condition. In the special case presented here, we consider the Hermite‐type weight functions and prove that the associated polynomial sequence can be expressed in the closed form via the Hermite polynomials. Also, we find their Rodrigues‐type formula and a four‐term recurrence relation. In contrast to the zeros of Hermite polynomials, which are symmetrically located with respect to the origin, the zeros of the new polynomial sequence are all positive. Copyright © 2015 John Wiley & Sons, Ltd.     

Linearization of the product of symmetric orthogonal polynomials

A new constructive approach is given to the linearization formulas of symmetric orthogonal polynomials. We use the monic three-term recurrence relation of an orthogonal polynomial system to set up a partial difference equation problem for the product of two polynomials and solve it in terms of the initial data. To this end, an auxiliary function of four integer variables is introduced, which may be seen as a discrete analogue of Riemann's function. As an application, we derive the linearization formulas for the associated Hermite polynomials and for their continuousq-analogues. The linearization coefficients are represented here in terms of₃ F ₂ and₃Φ₂ (basic) hypergeometric functions, respectively. We also give some partial results in the case of the associated continuousq-ultraspherical polynomials.