No label defined (Q3033239)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 4131569
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | No label defined |
scientific article; zbMATH DE number 4131569 |
Statements
1988
0 references
error bounds
0 references
p-order method
0 references
principal error equation
0 references
two-sided approximation methods
0 references
0.760372519493103
0 references
0.7536860108375549
0 references
[For the entire collection see Zbl 0627.00019.] With regard to the initial value problem (i) \(y'=f(x),\) \(y(x_ 0)=\eta,\) \(x\in [x_ 0,X]\), \(y,f\in R^ m\), discretization methods of the general form (ii) \(y_{n+1}=y_ n+h\Phi (x_ n,y_ n,y_{n+1};h)\) are considered. On the assumption that (ii) is a p-order method and its discretization error can expressed in the form \(E_ n=h^ pe(x_ n)+O(h^{p+1}),\) where e(x) satisfies the so called principal error equation (iii) \(e'(x)=f_ y(y)e(x)+\psi (y),\quad e(x_ 0)=0,\) some two-sided approximation methods for the problem (iii) are given. Such approximations allow for a bound of \(h^ pe(x)\), the principal part of \(E_ n\). Only the knowledge of a good approximation of y in (iii) is required: if p is large enough, such an approximation can be the same furnished by (ii) when applied to (i). A practical consequence of bounding e(x) is the possibility of improving the approximations \(y_ n\) given by (ii).
0 references