The repetition factor and numerical stability of volterra integral equations

McKee, Sean; Brunner, Hermann

doi:10.1016/0898-1221(80)90041-3

Cited by 9 publications

(3 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Linz describes qualitatively that behaviour which he considers unstable; Wolkenfelt [50 ] has proposed a compatible formal definition..(The first rigorous insight into this form of stability was provided by Kobayasi [31] ; cf. Noble [38], and [28], [35]. )…”

Section: Forms Of 1 Forms Of Casementioning

confidence: 99%

An introduction to the numerical treatment of volterra and abel-type integral equations

Baker

1982

Topics in Numerical Analysis

View full text Add to dashboard Cite

An integral equation can be described as a functional equation in which the unknown function appears as part of an integrand.The inadequacies of this definition will not concern us here, since our subject is the numerical solution of a subclass of integral equations which we shall specify below. We shall frequently make comparisons with topics in the treatment of ordinary differential equations, with which we assume the reader to be more familiar.In any specific endeavour, the numerical analyst should adopt a broad perspective. Ideally, he should relate the application area to the theoretical analysis,to the mathematical analysis of numerical schemes, and to the construction of algorithms and automatic software. Figure 1 illustrates schematically the interaction between the various areas.Some expansion on the meaning of some of the headings in Figure 1 is in order.Under analytic theory we include classical and functional analysis of the equations, the theory of ill-and well-posedness in various contexts, and inherent stability ("conditioning"), the asymptotic theory, and, for example, possible reformulation of the problem. What constitutes elegant theory may be a subjective assessment, but, frequently, the theory of discrete schemes is at its best when it mimics the theory of the functional equation.Under numerical methods we include the basic numerical techniques: primitive quadrature schemes, basic collocation methods etc. The theoretical numerical analysis has as afirst objective results on convergence, order, and numerical stability followed by the effect of rounding error. An important feature is an assessment of the ability to model, via the numerical scheme, relevant qualitative behaviour of the functional equations. (These considerations give rise to such notions as A-stability, for example.)We remark, in passing, that whereas many will take as a point of doctrine the necessity of convergence in a numerical scheme, as (say) some discretization parameter tends to zero, this property is not sufficient for practical computation. The modelling of important qualitative properties can be absent in practical computations with theoretically convergence schemes. An assessment of which qualitative properties are important can only result from familiarity with the area of application.The development of reliable and robust adaptive software is one aim of the Observe the possibility of confusion between the use of the word stability in the sense of the classical analyst (which relates to a form of what numerical analysts term inherent conditioning) and "numerical stability".

show abstract

Section: Forms Of 1 Forms Of Casementioning

confidence: 99%

An introduction to the numerical treatment of volterra and abel-type integral equations

Baker

1982

Topics in Numerical Analysis

View full text Add to dashboard Cite

show abstract

“…Clearly, we can expect stability in the sense of Noble, i.e., we will not have the catastrophic growth of spurious solutions introduced by the discretization since, in keeping with his analysis, we have a repetition factor of one (see [10] and [13]). On the other hand, we do not necessarily expect that the absolute error will remain bounded-in-norm, since that condition can be shown not to hold for a simpler version of (3.4), viz.…”

Section: Diagram 21mentioning

confidence: 99%

“…This scheme has been shown to be convergent and stable, even though it is unstable when the trapezoidal rule is applied on the left (see [10] and [13]). …”

Section: Diagram 21mentioning

confidence: 99%