Re: [python-control-discuss] ODE integrator problems

[Ryan K. <rk...@si...>]

I have run into this same problem with pulse inputs.  The issue is that the
underlying Fortran numeric integration algorithm tries to determine the
largest possible step-size that can be used and still get accurate results.
 If the input starts and ends at zero but is non-zero in the middle, it is
possible for the adaptive step-size algorithm to essentially step over the
input.  There is no way for the algorithm to correctly choose the correct
step-size in that situation without exhaustively trying every possible
point on the input vector.  So, setting hmax is a perfectly valid solution
in this case.  If you input also starts at zero, you could try truncating
the input vector to start at the first non-zero element.


--
Ryan Krauss, Ph.D.
Associate Professor
Mechanical Engineering
Southern Illinois University Edwardsville


On Tue, Mar 26, 2013 at 10:55 AM, Sawyer Fuller <mi...@se...>wrote:

> Is anybody experiencing terrible performance of the integrator, used in
> functions like "step_response"? I see glitches and other random behavior
> when running my own code, and even for example the pvtol-nested.py example
> code. The integrator it uses, scipy.integrate.odeint, is derived from some
> fortran code somewhere and  appears to be doing optimization that is giving
> unwanted behavior.
>
> One can force a short simulation interval dt for a fixed-step integration
> by passing "hmax=dt" to "step-respose" and get a response that looks like
> Matlab's, but I'd rather just have a working integrator that optimizes
> correctly.
>
> Anybody know a good workaround, eg settings for the integrator that make
> it behave normally? Or why it is giving odd behavior?
>
> S
>
>
> ------------------------
> Sawyer B. Fuller, Ph.D.
> Harvard School of Engineering and Applied Science
> Wyss Institute for Biologically Inspired Engineering
> http://people.seas.harvard.edu/~minster/
>
>
>
> ------------------------------------------------------------------------------
> Own the Future-Intel&reg; Level Up Game Demo Contest 2013
> Rise to greatness in Intel's independent game demo contest.
> Compete for recognition, cash, and the chance to get your game
> on Steam. $5K grand prize plus 10 genre and skill prizes.
> Submit your demo by 6/6/13. http://p.sf.net/sfu/intel_levelupd2d
>
> _______________________________________________
> python-control-discuss mailing list
> pyt...@li...
> https://lists.sourceforge.net/lists/listinfo/python-control-discuss
>
>