对流-扩散问题的特征──块中心差分法

１．引言１９８２年,Ｄｏｕｇｌａｓ和Ｒｕｓｓｅｌｌ［１］提出解对流一扩散问题的特征一差分方法,网格节点为均匀分布,求解区域为直线Ｒ．文中讨论了基于二次插值的特征一差分格式,但其近似解按离散Ｌ２模未达到最优阶误差估计．１９８８年Ｗｅｉｓｅｒ和Ｗｈｅｅｌｅｒ［２］提出解线性椭圆型和线性抛物型方程的块中心差分法,１９９１年王申林［３］讨论了解拟线性双曲型积分微分方程的块中心差分方法,其共同特点为近似解按离散的Ｌ２模达到最优阶误差估计,解的一阶导数的近似解达到超收敛误差估计．１９９３年由同顺［４］讨论了…     

截尾试验下指数分布的贝叶斯估计

在指数分布场合，定数或定时截尾试验，文［１］给出了参数λ在先验分布为Γ（α，β）分布的假设下的Ｂａｙｅｓ估计．文［３］给出了在平方损失下的Ｂａｙｅｓ估计．本文讨论先验分布为Ｂ（ａ，ｂ）分布时，参数λ的Ｂａｙｅｓ估计．     

修正的LMS方法

ＢｉｔｍｅａｄＲ．Ｒ和ＡｎｄｅｒｓｏｎＤ．Ｏ在文献［１］中为任意线性方程组的求解提出了一种颇为有效的算法，称为ＬＭＳ方法．文献［２］详细地论述了算法的收敛性，指出收敛极限是方程组的最小二乘解．本文为使解线性方程组的ＬＭＳ算法具有更广泛、更方便的应用性．对文献［２］中的ＬＭＳ算法作了修正．理论和实践证明修正后的算法是成功的．     

积分算子的n-逼近数

本文研究了积分算子ＴＫ：Ｌｑ［０，１］→Ｌｑ［０，１］，（ｑ≥１）当核 Ｋ（ｓ， ｔ）是 Ｓｏｂｏｌｅｖ空间 Ｗｐｒ（［０， １］２）中元素时ｎ－逼近数 ａｎ（ＴＫ： Ｌｑ→ Ｌｑ）的估计，并把这个估计应用于退化核方法解第二类线性Ｆｒｅｄｈｏｌｍ方程（Ｉ一ＴＫ）ｘ＝ｙ时，Ｂａｄｈｖａｌｏｖ［５］意义下最佳误差的讨论中，所得到的最佳误差之估计当ｑ＝１时，最优化了［１０］的结论．     

离散动力系统水动点类的Nielsen数

本文在［１］，［３］的基础上引入了离散动力系统不动点类的Ｎｉｅｌｓｅｎ－数，讨论了Ｎｉｅｌｓｅｎ－数的同伦性质，得到利用Ｎｉｅｌｅｓｎ－数估计离散动力系统不动点个数的几个结果，为进一步研究该系统不动点的个数问题提供了有用的工具。     

一反应扩散模型解的有限元方法和误差分析

１引言在地下水运移过程中,污染物（溶质）随地下水在含水层中运移,并常常发生各种化学反应．文献［１－３］等提出并论述了三种化学物质（如Ｍ１,Ｍ２和Ｍ３）之间发生的一类化学反应．文献［４,５］等建立和描述了这类反应的数学模型（Ｐ）．文献［６,７」的作者首次对模型（Ｐ）进行了理论上的定性分析,主要是利用上,下解方法,算子半群理论和Ｓｏｂｏｌｅｖ空间的般人定理等论证了模型（Ｐ）的整体古典解的存在唯一性和渐近性质,文［６,７］也讨论了整体解的极限性态和收敛性估计．此外,文［８,９］等也就模型（Ｐ）的一类特…     

γ一条件下 Hansen和 Patrick 方法的收敛性

１９７７年Ｈａｎｓｅｎ和Ｐａｔｒｉｃｋ提出了一族求复函数ｆ：Ｃ－－＋Ｃ零点的带参数Ａ的选代方法［２］在区间估计的判据下证明了此方法的收敛性;而［３］用Ｓｍａｌｅ的点估计判据证明了：当人Ｅ时,此方法对复解析函数是收敛的．但是解析性的条件太强了．［４］和...     

解高维广义对称正则长波方程的Fourier谱方法

１引言对称正则长波方程（ＳＲＬＷＥ）是正则化长波方程（ＲＬＷＥ）的一种对称叙述［１］用于描述弱非线性作用下空间变换的离子声波传播．［１］得到了方程组（１．１）的双曲正割平方孤立波解、四个不变量和数值结果、明显地,从（１．１）中消去ρ,得到一类正则长波方程（ＲＬＷＥ）代替（１．２）中第三项、第四项对ｔ的导数为对ｘ的导数,得到Ｂｏｕｓｓｉｎｅｓｑ方程．［２］对一类广义对称正则长波方程组提出了谱方法,证明了古典光滑解的存在性和唯一性,建立了近似解的收敛性和误差估计。［３］研究了高维对称正则长波方程整体…     

基于截断删失数据乘积限估计的一些逼近定理

本文证明了由Ｔｓａｉ等［２］所定义的基于截断删失数据乘积限估计可被一Ｕ—统计量以足够的精度逼近，以至由该Ｕ—统计量的ＢｅｒｙＥｓｅｎ定理可得此乘积限估计的ＢｅｒｙＥｓｓｅｎ不等式．此外，应用Ｇｉｂｊｅｌｓ等人［１，定理１（Ｃ）］的有关结果，得到了此处乘积限估计误差的ｒ（＞１）阶绝对矩不等式和泛函型重对数律     

线性模型参数估计的一种新的相对效率

本文对线性模型的最小二乘估计（ＬＳＥ）与最佳线性无偏估计（ＢＬＵＥ）提出了一种新的相对效率并给出了新的相对效率的上、下界,最后还讨论了新的相对效率与已有的几种相对效率的关系。     

On the Determination of a Density Function by Its Autoconvolution Coefficient

We investigate eigenvalues and eigenvectors of certain linear variational eigenvalue inequalities where the constraints are defined by a convex cone as in [4], [7], [8], [10]-[12], [17]. The eigenvalues of those eigenvalue problems are of interest in connection with bifurcation from the trivial solution of nonlinear variational inequalities. A rather far reaching theory is presented for the case that the cone is given by a finite number of linear inequalities, where an eigensolution corresponds to a (+)-Kuhn-Tucker point of the Rayleigh quotient. Application to an unlaterally supported beam are discussed and numerical results are given.     

Nonstandard Analysis Methods for Separations in[0,1]-topological Spaces

In nonstandard enlargement,the separations are characterized by nonstandard analysis methods in[0,l]-topological spaces.Firstly,the monads of fuzzy point in[0,l]-topological spaces are described with remote-neighborhoods in nonstandard enlarged model.Then the nonstandard characterizations of separations in [0,l]-topological space are given by the monads.At last,relations of these separations are investigated.     

Constructive methods for certain Bergman operators

This paper is concerned with the class of linear partial differential equations of second order such that there exist Bergman operators with polynomial kernels (cf, [12]). In an earlier paper [ll] the authors have shown that these equations also admit differential operators as introduced by K. W. Bauer [I]. In the present paper, relations between different types of representations of solutions are investigated. These representations are of interest in developing a function theory of solutions; cf., for instance, K. W. Bauer [I] and S. Ruscheweyh [19]. They are also essential to global extensions of local results obtained by means of Bergman operators of the first kind. The inversion problem for those operators is solved, and it is shown that all solutions of equations of that class which are holomorphic in a domain of C² can be represented by operators with polynomial kernels. Furthermore, a construction principle for deriving the equations investigated by K. W. Bauer [2] is obtained; this yields corresponding representations of solutions by differential and integral operators in a systematic fashion     

Higher holonomies: Comparing two constructions

We compare two different constructions of higher-dimensional parallel transport. On the one hand, there is the two-dimensional parallel transport associated with 2-connections on 2-bundles studied by Baez–Schreiber [2], Faria Martins–Picken [11] and Schreiber–Waldorf [12]. On the other hand, there are the higher holonomies associated with flat superconnections as studied by Igusa [7], Block–Smith [3] and Arias Abad–Schätz [1]. We first explain how by truncating the latter construction one obtains examples of the former. Then we prove that the two-dimensional holonomies provided by the two approaches coincide.     

Singuläre S-hermitesche Rand-Eigenwertprobleme

The theory of singular self-adjoint eigenvalue problems developed by Weyl [13], Stone [12], Kodaira [2] and others has been generalized by A. Schneider [8], [9], [10], [11] to real S-hermitian systems of differential equations with real boundary conditions. Here the theory of singular S-hermitian boundary-value problems for arbitrary complex systems of differential equations with complex boundary conditions is developed. Moreover the boundary conditions are allowed to depend linearly on the eigenvalue-parameter.     

A class of norms of iterative methods for solving systems of linear equations

A new class of norms which generalize norms previously investigated by Young [9, 14], Sheldon [4, 5], Golub [1], Golub and Varga [2], Varga [6], Wachspress [7], Young and Kincaid [12], Young [14], and Kincaid [3] is introduced. Expressions for these norms applied to the matrices associated with various iterative methods are developed.Work on this paper was sponsored by NSF Grant GP-8442 and Army Grant DA-ARO(D)-31-124-G1050 at The University of Texas at Austin.     

Concentration factors for functions with harmonic bounded mean variation

We discuss determination of jumps for functions with generalized bounded variation. The questions are motivated by A. Gelb and E. Tadmor [1], F. Móricz [5] and [6] and Q. L. Shi and X. L. Shi [7]. Corollary 1 improves the results proved in B. I. Golubov [2] and G. Kvernadze [3]. Supported by NSFC 10671062.     

Vacuum States and Equidistribution of the Random Sequence for Glimm's Scheme (continuation)

This is a continuation of paper [1]. The difference between this paper and paper [1] is that the initial functions considered here are step functions and those considered in [1] are. Lipschitz continuous. Since there are centered rarefaction waves here, more delicate techniques are needed. It may be a necessary step in solving p-System with general initial functions by Glimm's scheme. Notice that this paper can not be deduced from [1].     

Stabile Mehrschichtverfahren für parabolische Evolutionsgleichungen

Summary The line method discussed in [19] is improved by considering several time slices. It is applied to parabolic initial-value problems, and with techniques similar to [16] it is proved that the root conditions of Dahlquist [5] and Widlund [17] are sufficient for stability. Stable methods up to order 6 are given and illustrated by numerical calculations.

Die Arbeit ist im wesentlichen die gekürzte Fassung des zweiten Teils meiner am Lehrstuhl von Prof. G. Hämmerlin angefertigten Dissertation [18]. Über den anderen Teil wurde bereits in [19] berichtet     

矩阵多项式空间的进一步研究

设A为数域F上的n级矩阵,记F[A]={f(A)|f(x)∈F[x]},它显然是F~(n×n)的子空间.讨论了F[A]的基和维数,引入了f(A)的坐标和F[A]的因式子空间的概念,给出了用因式子空间表示F[A]的几个定理,刻画了F[A]的结构.