Local error estimates for moderately smooth problems: Part II—SDEs and SDAEs with small noise

The paper consists of two parts. In the first part of the paper, we proposed a procedure to estimate local errors of low order methods applied to solve initial value problems in ordinary differential equations (ODEs) and index-1 differential-algebraic equations (DAEs). Based on the idea of Defect Correction we developed local error estimates for the case when the problem data is only moderately smooth, which is typically the case in stochastic differential equations. In this second part, we will consider the estimation of local errors in context of mean-square convergent methods for stochastic differential equations (SDEs) with small noise and index-1 stochastic differential-algebraic equations (SDAEs). Numerical experiments illustrate the performance of the mesh adaptation based on the local error estimation developed in this paper. The first author acknowledges support by the BMBF-project 03RONAVN and the second author support by the Austrian Science Fund Project P17253.     

Reversible methods of Runge-Kutta type for Index-2 DAEs

Summary. A new interpretation of Runge-Kutta methods for differential algebraic equations (DAEs) of index 2 is presented, where a step of the method is described in terms of a smooth map (smooth also with respect to the stepsize). This leads to a better understanding of the convergence behavior of Runge-Kutta methods that are not stiffly accurate. In particular, our new framework allows for the unified study of two order-improving techniques for symmetric Runge-Kutta methods (namely post-projection and symmetric projection) specially suited for solving reversible index-2 DAEs.Mathematics Subject Classification (1991): 65L05, 65L06     

Linear Index-1 DAEs: Regular and Singular Problems

Several features and interrelations of projector methods and reduction techniques for the analysis of linear time-varying differential-algebraic equations (DAEs) are addressed in this work. The application of both procedures to regular index-1 problems is reviewed, leading to some new results which extend the scope of reduction techniques through a projector approach. Certain singular points are well accommodated by reduction methods; the projector framework is adapted in this paper to handle (not necessarily isolated) singularities in an index-1 context. The inherent problem can be described in terms of a scalarly implicit ODE with continuous operators, in which the leading coefficient function does not depend on the choice of projectors. The nice properties of projectors concerning smoothness assumptions are carried over to the singular setting. In analytic problems, the kind of singularity arising in the scalarly implicit inherent ODE is also proved independent of the choice of projectors. The discussion is driven by a simple example coming from electrical circuit theory. Higher index cases and index transitions are the scope of future research.     

Simulating digital circuits numerically – a charge-oriented ROW approach

A ROW type approach is considered for integral form DAEs arising in charge-oriented nodal analysis of digital networks. These network equations define very special index-2 systems that can be solved by Rosenbrock-Wanner (ROW) methods suitable for semi-explicit index-1 systems without order reduction. To obtain charge conservation, the charge variables are projected on the linear charge constraint. In contrast to the semi-explicit index-1 case, all order conditions for the algebraic variables up to order have to be fulfilled for a method of order . CHORAL, an embedded charge-oriented method of order (2)3, is introduced and compared with DASSL and RODAS for two industrial applications, the NAND gate and the two-bit adding unit. Received January 22, 1996 / Revised version received January 28, 1997     

A note on bifurcation points for index-1 differential-algebraic problems

The arclength continuation method is proposed for index-1 DAEs with singular points. In particular, the cusp points, the tangency points and equilibrium points are investigated. The numerical iterative matrix is studied at a singular point. The numerical examples are given to illustrate the robustness of the continuation method.     

Parallel collocation solution of index-1 BVP-DAEs arising from constrained optimal control problems

The indirect solution of constrained optimal control problems gives rise to two-point boundary value problems (BVPs) that involve index-1 differential-algebraic equations (DAEs) and inequality constraints. This paper presents a parallel collocation algorithm for the solution of these inequality constrained index-1 BVP-DAEs. The numerical algorithm is based on approximating the DAEs using piecewise polynomials on a nonuniform mesh. The collocation method is realized by requiring that the BVP-DAE be satisfied at Lobatto points within each interval of the mesh. A Newton interior-point method is used to solve the collocation equations, and maintain feasibility of the inequality constraints. The implementation of the algorithm involves: (i) parallel evaluation of the collocation equations; (ii) parallel evaluation of the system Jacobian; and (iii) parallel solution of a boarded almost block diagonal (BABD) system to obtain the Newton search direction. Numerical examples show that the parallel implementation provides significant speedup when compared to a sequential version of the algorithm.     

预测校正公式求解指标2微分代数方程

1．前言微分代数方程（EEES）是经常出现于实际问题中的一类方程．其数值求解已成为常微分方程数值求解领域十分活跃的一个方向．目前微分代数方程求解的数值方法主要是nunge－Kutta型方法及BDF方法．Runge－Kutta型方法在网,问中有详细的介绍．Hairer等人据此编制了软件RADAU,而目前使用最广泛的软件还是PetZold等编制的DASSL．DASSL使用的方法为BDF方法,它在微分代数方程中的应用最早可以追述到Gear的开创性工作问．BDF方法一个很大的优点是刚性稳定．然而对于非刚性的微分代数方程,刚性稳定已不是主要考虑的因素．因此…     

指数—2非线性广义系统的解及其稳定性

本文主要给出指数－2的非线性广义系统的可解性和零解的稳定性分析．     

Non-stiff integrators for differential–algebraic systems of index 2

Non-stiff differential-algebraic equations (DAEs) can be solved efficiently by partitioned methods that combine well-known non-stiff integrators from ODE theory with an implicit method to handle the algebraic part of the system. In the present paper we consider partitioned one-step and partitioned multi-step methods for index-2 DAEs in Hessenberg form and the application of these methods to constrained mechanical systems. The methods are presented from a unified point of view. The comparison of various classes of methods is completed by numerical tests for benchmark problems from the literature. This revised version was published online in June 2006 with corrections to the Cover Date.     

The computation of consistent initial values for nonlinear index-2 differential–algebraic equations

The computation of consistent initial values for differential–algebraic equations (DAEs) is essential for starting a numerical integration. Based on the tractability index concept a method is proposed to filter those equations of a system of index-2 DAEs, whose differentiation leads to an index reduction. The considered equation class covers Hessenberg-systems and the equations arising from the simulation of electrical networks by means of Modified Nodal Analysis (MNA). The index reduction provides a method for the computation of the consistent initial values. The realized algorithm is described and illustrated by examples.     

IRK Methods for Index 2 and 3 DAEs: Starting Algorithms

When semi-explicit differential-algebraic equations are solved with implicit Runge-Kutta methods, the computational effort is dominated by the cost of solving the non-linear systems. That is why it is important to have good starting values to begin the iterations. In this paper we study a type of starting algorithms, without additional computational cost, in the case of index-2 and index-3 DAEs. The order of the starting values is defined, and by using DA-series and rooted trees we obtain their general order conditions. If the RK method satisfies some simplifying assumptions, then the maximum order can be obtained.     

An efficient asymptotically correct error estimator for collocation solutions to singular index-1 DAEs

A computationally efficient a posteriori error estimator is introduced and analyzed for collocation solutions to linear index-1 DAEs (differential-algebraic equations) with properly stated leading term exhibiting a singularity of the first kind. The procedure is based on a modified defect correction principle, extending an established technique from the context of ordinary differential equations to the differential-algebraic case. Using recent convergence results for stiffly accurate collocation methods, we prove that the resulting error estimate is asymptotically correct. Numerical examples demonstrate the performance of this approach. To keep the presentation reasonably self-contained, some arguments from the literature on DAEs concerning the decoupling of the problem and its discretization, which is essential for our analysis, are also briefly reviewed. The appendix contains a remark about the interrelation between collocation and implicit Runge-Kutta methods for the DAE case.     

Structural analysis of electrical circuits including magnetoquasistatic devices

Modeling electric circuits that contain magnetoquasistatic (MQS) devices leads to a coupled system of differential-algebraic equations (DAEs). In our case, the MQS device is described by the eddy current problem being already discretized in space (via edge-elements). This yields a DAE with a properly stated leading term, which has to be solved in the time domain. We are interested in structural properties of this system, which are important for numerical integration. Applying a standard projection technique, we are able to deduce topological conditions such that the tractability index of the coupled problem does not exceed two. Although index-2, we can conclude that the numerical difficulties for this problem are not severe due to a linear dependency on index-2 variables.     

Runge-Kutta schemes that maintain invariant solutions for differential algebraic systems

Implicit Runge-Kutta (IRK) methods and projected IRK methods for the solution of semiexplicit index-2 systems of differential algebraic systems (DAEs) have been proposed by several authors. In this paper we prove that if a method satisfiesBA+A ^t B–bb ^t =0, it conserves quadratic invariants of DAEs.     

Trajectory sensitivity analysis of hybrid systems with sliding motion

This paper concerns hybrid control systems exhibiting the sliding motion. It is assumed that the system’s motion on the switching surface is described by index-2 differential–algebraic equations (DAEs), which guarantee the accurate tracking of the sliding motion surface. For those systems the sensitivity analysis is performed with the help of solutions to system’s linearized equations. The paper states conditions under which the solutions to the linearized equations for original DAEs and the solutions to linearized equations for underlying ordinary differential equations (ODEs) exhibit similar properties. Due to the presence of sliding motion, we restrict the class of admissible control functions to piecewise differentiable functions. The presented sensitivity analysis might be useful in deriving the weak maximum principle for optimal control problems with hybrid systems exhibiting sliding motion and in establishing the global convergence of algorithms for solving those problems.     

A multirate ROW-scheme for index-1 network equations

Multirate methods exploit latency in electrical circuits to simulate the transient behaviour more efficiently. To this end, different step-sizes are used for various subsystems. The size of these time steps reflect the different levels of activity. Following the idea of mixed multirate for ordinary differential equations, a Rosenbrock–Wanner based multirate method is developed for index-1 differential-algebraic equations (DAEs) arising in circuit simulation. To obtain order conditions for a method with two activity levels, P-series (and DA-series) are generalised and combined for an application to partitioned DAE systems. A working scheme and results for a benchmarking circuit are presented.     

Efficient Runge-Kutta integrators for index-2 differential algebraic equations

In seeking suitable Runge-Kutta methods for differential algebraic equations, we consider singly-implicit methods to which are appended diagonally-implicit stages. Methods of this type are either similar to those of Butcher and Cash or else allow for the importation of a final derivative from a previous step. For these two classes, with up to three additional diagonally-implicit stages, we derive methods that satisfy appropriate order conditions for index-2 DAEs.

     

Parameter estimation for DAE models in a multiple experiment context

Parameter estimation for PDE models is a recent topic of our research. After discretization in space we get a system of DAEs. We here present the techniques we use (e.g., Multiple Shooting, generalized Gauss-Newton) to fit the parameters. (© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim)     

One-step 9-stage Hermite–Birkhoff–Taylor DAE solver of order 10

The HBT(10)9 method for ODEs is expanded into HBT(10)9DAE for solving nonstiff and moderately stiff systems of fully implicit differential algebraic equations (DAEs) of arbitrarily high fixed index. A scheme to generate first-order derivatives at off-step points is combined with Pryce scheme which generates high order derivatives at step points. The stepsize is controlled by a local error estimator. HBT(10)9DAE uses only the first four derivatives of y instead of the first 10 required by Taylor’s series method T10DAE of order 10. Dormand–Prince’s DP(8,7)13M for ODEs is extended to DP(8,7)DAE for DAEs. HBT(10)9DAE wins over DP(8,7)DAE on several test problems on the basis of CPU time as a function of relative error at the end of the interval of integration. An index-5 problem is equally well solved by HBT(10)9DAE and T10DAE. On this problem, the error in the solution by DP(8,7)DAE increases as time increases.     

计算微分代数系统稳定域的迭代Lyapunov函数方法

用迭代Lyapunov函数方法对微分代数系统稳定域进行了研究,根据所研究的微分代数系统形式,构造一个Lyapunov函数,然后对这个Lyapunov函数进行逐次迭代,给出了微分代数系统稳定域逐次扩大的迭代算法,数值实验表明迭代Lyapunov函数方法应用于微分代数系统稳定域的估计比单个Lyapunov函数具有良好的优越性。