非定常Burgers方程基于POD方法的全二阶差分格式降阶外推算法

利用特征投影分解方法和奇值分解技术对非定常Burgers 方程的经典全二阶有限差分格式做降阶处理, 给出一种时间和空间变量都是二阶精度的降阶差分格式, 并给出这种降阶全二阶精度差分解的误差估计和外推算法的实现, 最后用数值例子说明数值结果与理论结果是相吻合的.     

守恒高阶各向异性交通流模型基于POD方法的降阶外推差分格式

利用Godunov流方法和特征投影分解方法,对守恒高阶各向异性交通流模型建立一种自由度很少、精度足够高的降阶外推差分算法, 并给出这种降阶外推差分算法近似解的误差估计和算法实现.最后,用数值例子说明数值结果与理论结果相吻合,并阐明这种降阶外推差分算法的优越性.     

二维土壤溶质输运方程基于POD方法的降阶CN有限体积元外推算法

利用Crank-Nicolson(CN)有限体积元方法和特征投影分解方法建立二维土壤溶质输运方程的一种维数很低、精度足够高的降阶CN有限体积元外推算法,并给出这种外推算法的降阶CN有限体积元解的误差估计和算法的实现.最后用数值例子说明数值结果与理论结果相吻合,并阐明这种降阶CN有限体积元外推算法的优越性.     

Sobolev方程基于POD的降阶外推差分算法

用奇异值分解和特征投影分解(proper orthogonal decomposition,简记POD)方法建立Sobolev方程的一种降阶外推有限差分算法,并给出误差估计.最后用数值例子,验证基于POD方法降阶外推有限差分算法的可行性和有效性.     

抛物方程基于POD方法的降阶外推差分算法

用奇值分解和特征投影分解(Proper Orthogonal Decomposition,简记POD)方法去建立抛物方程的一种降阶外推有限差分算法,并给出误差估计.最后用数值例子验证这种基于POD方法降阶外推有限差分算法的可行性和有效性.     

隐式重新启动的Lanczos算法在模型降阶中的应用

主要研究了一种隐式重新启动的L anczos算法在模型降阶中的应用.分析了由这个算法得到的降价后的模型的一些性质,对于一个n阶稳定的线性时不变系统,模型降阶的思想是寻找一个m阶转换函数来近似原系统的n阶转换函数H(s),其中n m.传统的kry lov子空间方法仅仅产生一个不稳定的实现,并且在低频处的误差较大,本文所考虑的隐式重新启动的L anczos方法,能较好的解决上述两个问题.     

二维土壤溶质输运方程基于POD方法的降阶CN有限元外推算法

首先给出二维土壤溶质输运方程时间二阶精度的Crank-Nicolson(CN)时间半离散化格式和时间二阶精度的全离散化CN有限元格式及其误差分析.然后利用特征投影分解(proper orthogonal decomposition,简记为POD)方法对二维土壤溶质输运方程的经典CN有限元格式做降阶处理,建立一种具有足够高精度、自由度很少的降阶CN有限元外推格式,并给出这种降阶CN有限元解的误差估计和外推算法的实现.最后用数值例子说明数值结果与理论结果是相吻合的.     

Hamilton系统下基于相位误差的精细辛算法

Hamilton系统是一类重要的动力系统，辛算法（如生成函数法、SRK法、SPRK法、多步法等）是针对Hamilton系统所设计的具有保持相空间辛结构不变或保Hamilton函数不变的算法.但是，时域上，同阶的辛算法与Runge-Kutta法具有相同的数值精度，即辛算法在计算过程中也存在相位误差，导致时域上解的数值精度不高.经过长时间计算后，计算结果在时域上也会变得“面目全非”.为了提高辛算法在时域上解的精度，将精细算法引入到辛差分格式中，提出了基于相位误差的精细辛算法（HPD-symplectic method），这种算法满足辛格式的要求，因此在离散过程中具有保Hamilton系统辛结构的优良特性.同时，由于精细化时间步长，极大地减小了辛算法的相位误差，大幅度提高了时域上解的数值精度，几乎可以达到计算机的精度，误差为O（10-13）.对于高低混频系统和刚性系统，常规的辛算法很难在较大的步长下同时实现对高低频精确仿真，精细辛算法通过精细计算时间步长，在大步长情况下，没有额外增加计算量，实现了高低混频的精确仿真.数值结果验证了此方法的有效性和可靠性.     

非饱和土壤水流方程基于特征投影分解方法的降阶CN有限元外推模型

利用Crank-Nicolson有限元方法和特征投影分解方法去建立二维非饱和土壤水流方程的一种维数很低,精度足够高的降阶CN有限元外推模型,并给出这种降阶CN有限元外推模型的降阶近似解误差估计和算法实现.最后用数值例子说明数值结果与理论结果相吻合,并阐明这种降阶CN有限元外推模型的优越性.     

二维双曲方程基于POD方法的降阶有限差分外推迭代格式

利用特征投影分解(POD)方法建立二维双曲型方程的一种基于POD方法的含有很少自由度但具有足够高精度的降阶有限差分外推迭代格式,给出其基于POD方法的降阶有限差分解的误差估计及基于POD方法的降阶有限差分外推迭代格式的算法实现.用一个数值例子去说明数值计算结果与理论结果相吻合.进一步说明这种基于POD方法的降阶有限差分外推迭代格式对于求解二维双曲方程是可行和有效的.     

Lebesgue piecewise affine approximation of nonlinear systems

This paper addresses a piecewise affine (PWA) approximation problem, i.e., a problem of finding a PWA system model which approximates a given nonlinear system. First, we propose a new class of PWA systems, called the Lebesgue PWA approximation systems, as a model to approximate nonlinear systems. Next, we derive an error bound of the PWA approximation model, and provide a technique for constructing the approximation model with specified accuracy. Finally, the proposed method is applied to a gene regulatory network with nonlinear dynamics, which shows that the method is a useful approximation tool.     

Second Order Asymptotics of Aggregated Log-Elliptical Risk

In this paper we establish the error rate of first order asymptotic approximation for the tail probability of sums of log-elliptical risks. Our approach is motivated by extreme value theory which allows us to impose only some weak asymptotic conditions satisfied in particular by log-normal risks. Given the wide range of applications of the log-normal model in finance and insurance our result is of interest for both rare-event simulations and numerical calculations. We present numerical examples which illustrate that the second order approximation derived in this paper significantly improves over the first order approximation.     

Rules for selecting the parameters of Oustaloup recursive approximation for the simulation of linear feedback systems containing PID controller

Oustaloup recursive approximation (ORA) is widely used to find a rational integer-order approximation for fractional-order integrators and differentiators of the form s^v, v ∈ (−1, 1). In this method the lower bound, the upper bound and the order of approximation should be determined beforehand, which is currently performed by trial and error and may be inefficient in some cases. The aim of this paper is to provide efficient rules for determining the suitable value of these parameters when a fractional-order PID controller is used in a stable linear feedback system. Two numerical examples are also presented to confirm the effectiveness of the proposed formulas.     

Error bounds for the large time step Glimm scheme applied to scalar conservation laws

Summary. In this paper we derive an error bound for the large time step, i.e. large Courant number, version of the Glimm scheme when used for the approximation of solutions to a genuinely nonlinear, i.e. convex, scalar conservation law for a generic class of piecewise constant data. We show that the error is bounded by for Courant numbers up to 1. The order of the error is the same as that given by Hoff and Smoller [5] in 1985 for the Glimm scheme under the restriction of Courant numbers up to 1/2. Received April 10, 2000 / Revised version received January 16, 2001 / Published online September 19, 2001     

Passivity-preserving model reduction of differential-algebraic equations in circuit simulation

We present an extension of the positive real balanced truncation model reduction method for differential-algebraic equations that arise in circuit simulation. This method is based on balancing the solutions of the projected generalized algebraic Riccati equations. Important properties of this method are that passivity is preserved in the reduced-order model and that there exists an approximation error bound. Numerical solution of the projected Riccati equations using the special structure of circuit equations is also discussed. (© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)     

On the numerical approach of the enthalpy method for the Stefan problem

In this article an error bound is derived for a piecewise linear finite element approximation of an enthalpy formulation of the Stefan problem; we have analyzed a semidiscrete Galerkin approximation and completely discrete scheme based on the backward Euler method and a linearized scheme is given and its convergence is also proved. A second‐order error estimates are derived for the Crank‐Nicolson Galerkin method. In the second part, a new class of finite difference schemes is proposed. Our approach is to introduce a new variable and transform the given equation into an equivalent system of equations. Then, we prove that the difference scheme is second order convergent. © 2004 Wiley Periodicals, Inc. Numer Methods Partial Differential Eq, 2004     

Superconvergence of conforming finite element for fourth‐order singularly perturbed problems of reaction diffusion type in 1D

We consider conforming finite element approximation of fourth‐order singularly perturbed problems of reaction diffusion type. We prove superconvergence of standard C¹ finite element method of degree p on a modified Shishkin mesh. In particular, a superconvergence error bound of in a discrete energy norm is established. The error bound is uniformly valid with respect to the singular perturbation parameter ?. Numerical tests indicate that the error estimate is sharp. © 2013 Wiley Periodicals, Inc. Numer Methods Partial Differential Eq 30: 550–566, 2014     

Computing the Approximation Error for Neural Networks with Weights Varying on Fixed Directions

We obtain a sharp lower bound estimate for the approximation error of a continuous function by single hidden layer neural networks with a continuous activation function and weights varying on two fixed directions. We show that for a certain class of activation functions this lower bound estimate turns into equality. The obtained result provides us with a method for direct computation of the approximation error. As an application, we give a formula, which can be used to compute instantly the approximation error for a class of functions having second order partial derivatives.     

Risk bounds for model selection via penalization

Performance bounds for criteria for model selection are developed using recent theory for sieves. The model selection criteria are based on an empirical loss or contrast function with an added penalty term motivated by empirical process theory and roughly proportional to the number of parameters needed to describe the model divided by the number of observations. Most of our examples involve density or regression estimation settings and we focus on the problem of estimating the unknown density or regression function. We show that the quadratic risk of the minimum penalized empirical contrast estimator is bounded by an index of the accuracy of the sieve. This accuracy index quantifies the trade-off among the candidate models between the approximation error and parameter dimension relative to sample size. If we choose a list of models which exhibit good approximation properties with respect to different classes of smoothness, the estimator can be simultaneously minimax rate optimal in each of those classes. This is what is usually called adaptation. The type of classes of smoothness in which one gets adaptation depends heavily on the list of models. If too many models are involved in order to get accurate approximation of many wide classes of functions simultaneously, it may happen that the estimator is only approximately adaptive (typically up to a slowly varying function of the sample size). We shall provide various illustrations of our method such as penalized maximum likelihood, projection or least squares estimation. The models will involve commonly used finite dimensional expansions such as piecewise polynomials with fixed or variable knots, trigonometric polynomials, wavelets, neural nets and related nonlinear expansions defined by superposition of ridge functions. Received: 7 July 1995 / Revised version: 1 November 1997     

Error Bound for Bernstein-Bézier Triangular Approximation

Based upon a new error bound for the linear interpolant to a function defined on a triangle and having continuous partial derivatives of second order, the related error bound for n-th Bernstein triangular approximation is obtained. The order of approximation is 1/n.