In Infinity the following
variables are used: :
1.
- the right border of the calculation interval
a default value - 10;
possible values - SupT
> 0;
2.
-a local error (the top level of the absolute local error limit) in
calculation.
a default value - 1e-5;
possible values -
> 0;
3.
- a system complexity flag (a system with specific features)
a default value - 1 (a system with specific
features);
possible values -
:= 1 or
:= 0;
4.
- accuracy: the number of retained signs after a comma in a numerical
work.
possible values -
:= 5..4*106;
At the calculation accuracy increase the time of calculation increases
as well;
5.
- an independent integration variable.
Note :
a )
:= 1 is a more "rigid" condition
of solution searching. In this case you have a 100 % guarantee that
the precise solution of any problems is inside the found area, but
due to the "rigidity" increase the calculation’s computing
expenses increase as well. If
:= 0, then the precise solution is inside
the found area only for the tasks which don’t possess any specific
features (without the second kind discontinuities or locally unstable
areas).
b ) The less the
variable value, the narrower the area containing the precise solution.
But if the problem possesses some specific features (for instance,
there is a discontinuity of the second kind in the solution –
Example: A problem with several discontinuances
of the second kind, or one of the solutions is locally unstable
- Example: A nonlinear system with a locally
unstable area), the area containing the precise solution is extended
because of the cumulative error (the top rating of a limiting absolute
full error). With the help of the described variables it is possible
to operate the calculation procedure of the entered ODE system.
Copyright
© 1998-2004, MathRevolt.
All rights reserved
|