Series expansion of weighted pseudoinverse matrices and iterative methods for calculating weighted pseudoinverse matrices and weighted normal pseudosolutions

Weighted pseudoinverse matrices are expanded into matrix power series with negative exponents and arbitrary positive parameters. Based on this expansion, iterative methods for evaluating weighted pseudoinverse matrices and weighted normal pseudosolutions are designed and analyzed. The iterative methods for weighted normal pseudosolutions are extended to solving constrained least-squares problems. __________ Translated from Kibernetika i Sistemnyi Analiz, No. 1, pp. 32–62, January–February 2006.     

Constructing third-order derivative-free iterative methods

In this work, we develop nine derivative-free families of iterative methods from the three well-known classical methods: Chebyshev, Halley and Euler iterative methods. Methods of the developed families consist of two steps and they are totally free of derivatives. Convergence analysis shows that the methods of these families are cubically convergent, which is also verified through the computational work. Apart from being totally free of derivatives, numerical comparison demonstrates that the developed methods perform better than the three classical methods.     

Parallel iterative methods for parabolic equations

For the problems of the parabolic equations in one- and two-dimensional space, the parallel iterative methods are presented to solve the fully implicit difference schemes. The methods presented are based on the idea of domain decomposition in which we divide the linear system of equations into some non-overlapping sub-systems, which are easy to solve in different processors at the same time. The iterative value is proved to be convergent to the difference solution resulted from the implicit difference schemes. Numerical experiments for both one- and two-dimensional problems show that the methods are convergent and may reach the linear speed-up.     

Generalized newton multi-step iterative methods GMNp,m for solving system of nonlinear equations

A generalization of the Newton multi-step iterative method is presented, in the form of distinct families of methods depending on proper parameters. The proposed generalization of the Newton multi-step consists of two parts, namely the base method and the multi-step part. The multi-step part requires a single evaluation of function per step. During the multi-step phase, we have to solve systems of linear equations whose coefficient matrix is the Jacobian evaluated at the initial guess. The direct inversion of the Jacobian it is an expensive operation, and hence, for moderately large systems, the lower-upper triangular factorization (LU) is a reasonable choice. Once we have the LU factors of the Jacobian, starting from the base method, we only solve systems of lower and upper triangular matrices that are in fact computationally economical. The developed families involve unknown parameters, and we are interested in setting them with the goal of maximizing the convergence order of the global method. Few families are investigated in some detail. The validity and numerical accuracy of the solution of the system of nonlinear equations are presented via numerical simulations, also involving examples coming from standard approximations of ordinary differential and partial differential nonlinear equations. The obtained results show the efficiency of constructed iterative methods, under the assumption of smoothness of the nonlinear function.     

三维对流扩散方程的稀疏存储及预条件迭代

基于四阶紧致格式对三维对流扩散方程进行离散,并给出所得到的离散线性方程组的块三角稀疏矩阵形式。以带双阈值的不完全因子化LU分解[(ILUT(τ,s))]作为预条件子,分别用FGMRES、BICGSTAB和TFQMR作为迭代加速器,对离散线性方程组进行求解验证了格式精度并比较了不同迭代法的CPU时间和迭代步。此外,通过比较传统迭代法和预条件迭代法的计算效率,表明预条件迭代法不仅能够保证格式的四阶精度,还能极大地提高收敛效率。     

The solution of the poisson-boltzmann equation between two spheres: Modified iterative method

We present a modification on the successive overrelaxation (SOR) method and the iteration of the Green's function integral representation for the solution of the (nonlinear) Poisson-Boltzmann equation between two spheres. In comparison with other attempts, which approximate the geometry or the nonlinearity, the computations here are done for the full problem and compared with those done by the finite element method as a typical method for such problems. For the parameters of general interest, while the SOR method does not work, and the iteration of the integral representation is limited in its convergence, our modification to these iterative schemes converge. The modified SOR surpasses both methods in simplicity and speed; it is about 100 times faster than the modified iteration of the integral representation, with the latter being still simpler and faster than the finite element method. These two examples further illustrate the advantage of our recent modification to iterative methods, which is based on an analytical fixed point argument.     

Multipoint efficient iterative methods and the dynamics of Ostrowski's method

ABSTRACT

In this work, we introduce a modification into the technique, presented in A. Cordero, J.L. Hueso, E. Martínez, and J.R. Torregrosa [Increasing the convergence order of an iterative method for nonlinear systems, Appl. Math. Lett. 25 (2012), pp. 2369–2374], that increases by two units the convergence order of an iterative method. The main idea is to compose a given iterative method of order p with a modification of Newton's method that introduces just one evaluation of the function, obtaining a new method of order p+2, avoiding the need to compute more than one derivative, so we improve the efficiency index in the scalar case. This procedure can be repeated n times, with the same approximation to the derivative, obtaining new iterative methods of order p+2n. We perform different numerical tests that confirm the theoretical results. By applying this procedure to Newton's method one obtains the well known fourth order Ostrowski's method. We finally analyse its dynamical behaviour on second and third degree real polynomials.     

On the dynamics of some Newton's type iterative functions

The dynamics of a family of Newton's type iterative methods for second- and third-degree complex polynomials is studied. The conjugacy classes of these methods are presented. Classical properties of rational maps are used.     

Fourth-order two-step iterative methods for determining multiple zeros of non-linear equations

In this paper, we describe and analyse two two-step iterative methods for finding multiple zeros of non-linear equations. We prove that the methods have fourth-order convergence. The methods calculate the multiple zeros with high accuracy. These are the first two-step multiple zero finding methods. The numerical tests show their better performance in the case of algebraic as well as non-algebraic equations     

非线性重复运动系统的双迭代优化学习控制

本文针对一类具有非参数不确定性和输出约束的非线性系统,提出一种双迭代优化学习控制策略,将复杂的迭代学习过程简化为两个相对简单的迭代控制器.首先引入一类饱和非线性函数不仅可以满足系统的位置约束,同时能够保证系统跟踪误差收敛于给定的邻域,然后针对每次迭代初始误差设计参考轨迹自修正策略,在每个迭代周期上设置一个固定的调整时间域,根据上次迭代的输出调整下一次迭代的参考轨迹.双迭代的控制结构可以同时更新两个迭代控制器的参数,来处理系统的非参数不确定性.进一步利用Barrier复合能量函数证明双迭代控制策略的收敛性和稳定性,并给出收敛条件.最后,通过一个算例证明了该控制策略的有效性.     

Singular perturbation and iterative separation of time scales

This tutorial paper presents an iterative method for the separation of slow and fast modes, which removes the inconsistencies of the classical quasi-steady-state approach and systematically improves the accuracy of the lower order models. It also serves as a self-contained introduction to singular perturbations. State variable reformulation and time scale identification are discussed and illustrated with power system examples. A correction procedure for nonlinear systems is also presented.     

Block iterative methods for the numerical solution of three dimensional mildly non-linear biharmonic problems of first kind

In this article, we discuss two sets of new finite difference methods of order two and four using 19 and 27 grid points, respectively over a cubic domain for solving the three dimensional nonlinear elliptic biharmonic problems of first kind. For both the cases we use block iterative methods and a single computational cell. The numerical solution of (?u/?n) are obtained as by-product of the methods and we do not require fictitious points in order to approximate the boundary conditions. The resulting matrix system is solved by the block iterative method using a tri-diagonal solver. In numerical experiments the proposed methods are compared with the exact solutions both in singular and non-singular cases.     

On a third order family of methods for solving nonlinear equations

In this article we present a third-order family of methods for solving nonlinear equations. Some well-known methods belong to our family, for example Halley's method, method (24) from [M. Basto, V. Semiao, and F.L. Calheiros, A new iterative method to compute nonlinear equations, Appl. Math. Comput. 173 (2006), pp. 468–483] and the super-Halley method from [J.M. Gutierrez and M.A. Hernandez, An acceleration of Newton's method: Super-Halley method, Appl. Math. Comput. 117 (2001), pp. 223–239]. The convergence analysis shows the third order of our family. We also give sufficient conditions for the stopping inequality |x _n+1?α|≤|x _n+1?x _n | for this family. Comparison of the family members shows that there are no significant differences between them. Several examples are presented and compared.     

Convergence of P-regular splitting iterative methods for non-Hermitian positive semidefinite linear systems

In this paper, we study the convergence of P-regular splitting iterative methods for singular and non-singular non-Hermitian positive semidefinite linear systems, which generalize the known results. Some numerical experiments are performed to illustrate the convergence results.     

Flyback CCM inverter for AC module applications: iterative learning control and convergence analysis

This paper presents an iterative learning controller (ILC) for an interleaved flyback inverter operating in continuous conduction mode (CCM). The flyback CCM inverter features small output ripple current, high efficiency, and low cost, and hence it is well suited for photovoltaic power applications. However, it exhibits the non-minimum phase behaviour, because its transfer function from control duty to output current has the right-half-plane (RHP) zero. Moreover, the flyback CCM inverter suffers from the time-varying grid voltage disturbance. Thus, conventional control scheme results in inaccurate output tracking. To overcome these problems, the ILC is first developed and applied to the flyback inverter operating in CCM. The ILC makes use of both predictive and current learning terms which help the system output to converge to the reference trajectory. We take into account the nonlinear averaged model and use it to construct the proposed controller. It is proven that the system output globally converges to the reference trajectory in the absence of state disturbances, output noises, or initial state errors. Numerical simulations are performed to validate the proposed control scheme, and experiments using 400-W AC module prototype are carried out to demonstrate its practical feasibility.     

有理模型辨识的两类新方法——混合迭代与柔性最小二乘法

针对有理模型提出两类辨识方法.首先提出基于递阶辨识思想的混合辨识方法,将模型分解为分子和分母两个子模型,分别用最小二乘法辨识分子参数,用粒子群算法和智能多步长梯度迭代算法辨识分母参数.由于降低了模型维数,且信息向量与噪声不相关,相对于传统的偏差补偿最小二乘算法,混合迭代法可以提高辨识精度并降低计算量.然后,为消除模型结构已知的假设,且充分利用最新数据更新系统参数,提出柔性递推最小二乘辨识方法,将有理模型转化为时变参数系统,进而辨识出时变系统的参数.仿真例子验证了所提出方法的有效性.     

Parallel,iterative solution of sparse linear systems: Models and architectures

Solving large, sparse, linear systems of equations is a fundamental problems in large scale scientific and engineering computation. A model of a general class of asynchronous, iterative solution methods for linear systems is developed. In the model, the system is solved by creating several cooperating tasks that each compute a portion of the solution vector. A data transfer model predicting both the probability that data must be transferred between two tasks and the amount of data to be transferred is presented. This model is used to derive an execution time model for predicting parallel execution time and an optimal number of tasks given the dimension and sparsity of the coefficient matrix and the costs of computation, synchronization, and communication.The suitability of different parallel architectures for solving randomly sparse linear systems is discussed. Based on the complexity of task scheduling, one parallel architecture, based on a broadcast bus, is presented and analyzed.     

Representations and expansions of weighted pseudoinverse matrices, iterative methods, and problem regularization. I. Positive definite weights

The paper reviews studies on the representations and expansions of weighted pseudoinverse matrices with positive definite weights and on iterative methods and regularized problems for calculation of weighted pseudoinverse matrices and weighted normal pseudosolutions. The use of these methods to solve constrained least-squares problems is examined. __________ Translated from Kibernetika i Sistemnyi Analiz, No. 1, pp. 47–73, January–February 2008.     

Representations and expansions of weighted pseudoinverse matrices, iterative methods, and problem regularization. II. Singular weights

The paper reviews studies on the representations and expansions of weighted pseudoinverse matrices with positive semidefinite weights and on the construction of iterative methods and regularized problems for the calculation of weighted pseudoinverses and weighted normal pseudosolutions based on these representations and expansions. The use of these methods to solve constrained least squares problems is examined. Continued from Cybernetics and Systems Analysis, 44, No. 1, 36–55 (2008). __________ Translated from Kibernetika i Sistemnyi Analiz, No. 3, pp. 75–102, May–June 2008.