Panel模型中两步估计的优良性

本文研究Panel模型中未知参数的估计问题，给出了两步估计的协方差的准确表达式．用均方误差作为度量估计的优劣标准，我们建立了两步估计优于Within估计和最小二乘估计的充要条件．特别我们获得了两步估计优于Within估计的简单充分条件．一般说来，对于中等数量的样本容量，两步估计就优于Within估计，类似的结论对Between估计或最小二乘估计也成立．     

对称熵损失下成功概率P的E—Bayes估计

研究对称熵损失下成功概率P的Bayes估计和E—Bayes估计,证明了前者的存在性及唯一性．模拟结果表明E-Bayes估计优于极大似然估计和Bayes估计．并将E—Bayes方法应用在证券投资预测之中,预测效果较好．     

具有不同光滑变量的变系数模型

本文探讨具有不同光滑变量的变系数模型的建模、估计和估计的渐近性．首先,从实际出发建立模型;然后,使用局部线性方法给出模型中未知函数的初始估计,再使用平均方法,给出它们的平均估计;进—步,给出这些平均估计的渐近正态性．两个模拟例子说明这一估计方法是有效的．     

强相依数据的函数系数部分线性模型的估计

本文讨论在数据是强相依的情况下函数系数部分线性模型的估计．首先,采用局部线性方法,给出该模型函数项函数的估计;然后,使用两阶段方法给出系数函数的估计．并且讨论了函数项函数估计的渐近正态性,以及系数函数估计的弱相合性和渐近正态性.模拟研究显示,这些估计是较为理想的．     

广义偏最小二乘(PLS)估计类

本文研究了一类含有偏最小二乘（partialleastsquaresPLS）估计的估计类．给出了PLS估计的一般代数形式；讨论了含有PLS估计的广义PPLS估计类的统计性质；给出了该估计类优于最小二乘估计的条件．     

强相关数据回归函数的小波估计

讨论了在强相关数据情形下对回归函数的小波估计,并且给出了估计量的均方误差的一个渐近展开表示式． 对研究估计量的优劣,所推导的近似表示式显得非常重要．对一般的回归函数核估计,如果回归函数不是充分光滑,这个均方误差表示式并不成立A·D2但对小波估计,即使回归函数间断连续,这个均方误差表示式仍然成立．因此,小波估计的收敛速度要比核估计来得快,从而小波估计在某种程度上改进了现有的核估计．     

误差方差的非齐次估计的可容许的充要条件

研究误差方差的非齐次二次估计的可容许性．在平方损失下,给出了一个非齐次二次估计在非齐次二次估计类中是误差方差的容许估计的充要条件．     

部分线性回归模型的M-估计

本文讨论部分线性回归模型的M-估计．用局部线性方法给出未知函数的M-估计，用两步估计方法给出参数的M-估计．进一步证明了未知函数的M-估计的弱一致性和渐近正态性，参数的M-估计的弱一致性．     

多元极值分布参数的最大似然估计与分步估计

本文考虑多元极值分布的参数估计，给出了分步估计渐近协差阵的近似表示，并对维数P=2，5及相关参数α=0．001，0．01；0．1（0．2），0．9；0．99，0．999的各种组合，计算了分步估计关于最大似然估计的渐近效率，分析了各种参数及维数对渐近效率的影响．分步估计是一种合理、简单而且有较强实用意义的估计方法．     

Bayes估计的进一步研究及其Pitman最优性

对线性模型参数,讨论了Bayes估计的Pitman最优性,将已有结果进行了改进,去掉了附加条件,证明了在Pitman准则下,Bayes估计一致优于最小二乘估计(LSE),在此基础上,提出了一种基于先验信息的方差分量估计,通过和基于LSE的方差分量估计作比较,证明了新估计是无偏估计且有更小的均方误差．最后,证明了在Pitman准则下生长曲线模型参数的Bayes估计优于最佳线性无偏估计．     

THE UPWIND OPERATOR SPLITTING FINITE DIFFERENCE METHOD FOR COMPRESSIBLE TWO-PHASE DISPLACEMENT PROBLEM AND ANALYSIS

For compressible two-phase displacement problem, a kind of upwind operator splitting finite difference schemes is put forward and make use of operator splitting, of calculus of variations, multiplicative commutation rule of difference operators, decompo-sition of high order difference operators and prior estimates are adopted. Optimal order estimates in L2 norm are derived to determine the error in the approximate solution.     

三维多组分可压缩驱动问题的分数步特征差分方法

本文提出三维多组分驱动问题的分数步长特征差分格式,并应用变分形式,能量方法,粗细网格配套,叁二次插值,高阶差分算子的分解和乘积交换性理论和技巧,得到最佳阶L^2误差估计,该方法已成功应用到油资源评估,强化采油数值模拟和海水入侵预测和防治的数值模拟中。     

非线性多层渗流系统的数值方法及其应用

对多层非线性渗流系统提出适合并行计算的迎风分数步差分格式，利用变分形式、能量方法、差分算子乘积交换性、高阶差分算子的分解、先验估计的理论和技巧，得到收敛性的最佳阶的误差估计．该方法已成功的应用到油资源渗流力学运移聚集数值模拟的生产实际中．     

GENERALIZED DIFFERENCE METHODS ON ARBITRARY QUADRILATERAL NETWORKS

1.IntroductionConsidertheboundaryvalueproblemofthePoissonequationwherefiisaconvexpolygonregon;r~OfftheboundaryofOandf(x,y)aknownfunctiononfi.Thegeneralizeddifferencemethodsonquadrilateralnetworksforellipticequationsareproposedin[11],'wheretheconverge...     

FINITE DIFFERENCE FRACTIONAL STEP METHODS FOR THE TRANSIENT BEHAVIOR OF A SEMICONDUCTOR DEVICE

Characteristic finite difference fractional step schemes are put forward. The electric potential equation is described by a seven-point finite difference scheme, and the electron and hole concentration equations are treated by a kind of characteristic finite difference fractional step methods. The temperature equation is described by a fractional step method. Thick and thin grids are made use of to form a complete set. Piecewise threefold quadratic interpolation, symmetrical extension, calculus of variations, commutativity of operator product, decomposition of high order difference operators and prior estimates are also made use of. Optimal order estimates in l2 norm are derived to determine the error of the approximate solution. The well-known problem is thorongley and completely solred.     

The Modified Upwind Finite Difference Fractional Steps Method for Compressible Two-phase Displacement Problem

For compressible two-phase displacement problem,the modified upwind finite difference fractionalsteps schemes are put forward.Some techniques,such as calculus of variations,commutative law of multiplicationof difference operators,decomposition of high order difference operators,the theory of prior estimates and tech-niques are used.Optimal order estimates in L~2 norm are derived for the error in the approximate solution.Thismethod has already been applied to the numerical simulation of seawater intrusion and migration-accumulationof oil resources.     

AnL 2-error estimate for an approximation of the solution of a parabolic variational inequality

Summary We estimate the order of the difference between the numerical approximation and the solution of a parabolic variational inequality. The numerical approximation is obtained using a finite element discretization in space and a finite difference discretization in time which is more general than is used in the literature. We obtain better error estimates than those given in the literature. The error estimates are compared with numerical experiments.     

The upwind finite difference fractional steps method for combinatorial system of dynamics of fluids in porous media and its application

For combinatorial system of multilayer dynamics of fluids in porous media, the second order and first order upwind finite difference fractional steps schemes applicable to parallel arithmetic are put forward and two-dimensional and three-dimensional schemes are used to form a complete set. Some techniques, such as implicit-explicit difference scheme, calculus of variations, multiplicative commutation rule of difference operators, decomposition of high order difference operators and prior estimates, are adopted. Optimal order estimates in L ² norm are derived to determine the error in the second order approximate solution. This method has already been applied to the numerical simulation of migration-accumulation of oil resources. Keywords: combinatorial system, multilayer dynamics of fluids in porous media, two-class upwind finite difference fractional steps method, convergence, numerical simulation of energy sources.     

非线性渗流耦合系统的数值方法及其应用

对非线性二相渗流耦合系统提出修正迎风分数步差分格式,利用变分形式、归纳法假定、高阶差分算子的分解、微分方程的先验估计理论和技巧,得到收敛性的最佳阶误差估计．该方法已成功应用于防治海水入侵主要工程后效预测与调控模式的生产实践中．     

三维非线性对流扩散问题的数值方法在渗流力学的应用

对三维非线性对流扩散问题提出一类适合并行计算的二阶迎风分数步差分格式,采用分数步技术,将三维问题化为连续解3个一维问题计算．利用变分形式、能量方法、差分算子乘积交换性、高阶差分算子的分解、微分方程先验估计的理论和技巧,得到收敛性的最佳阶的误差估计．该方法已成功的应用油资源运移聚集渗流力学数值模拟计算、海水入侵预测和防治的工程实践中．