Nice to have found this forum, it might be very useful.
I am on a leg press fitness case at this moment and I am trying out hte
paramter studies, it works fine, but my results are very jumpy. I was
wondering what the reason of this phenomenon could be. I attached a
picture in ‘files’ that shows this effect. I am only varying seat
angle. The question is, whether these types of sharp changes are of
mathematical errors or biomechanical. The amount of steps is 40, and
which is high enough to prevent effects of too little steps. Could you
explain to me, why the results of the parameterstudies are so jumpy and
not some more smoothly divided?
I am not sure what your output from the parameter study is?. Is it
the muscle activities, joint reactions or something else that gives
the jumpy results?
If the model is driven by mocap data this may introduce “noise” to
the model in terms of flutacting accelerations if the data are not
filtered sufficiently.
If the fluctuations are seen as “spikes” in the muscle activity you
could try add a small linear penalty to the solver it may remove
such “spikes”.
Normally parameter surfaces are relatively smooth, maybee you should
check the model manually for two settings in your parameter study,
in order to explain the difference.
I hope this helps you move on
Best regards
Søren, AnyBody Support
— In anyscript@yahoogroups.com, “pamela” <peemonline@…> wrote:
>
> Hi all,
>
> Nice to have found this forum, it might be very useful.
>
> I am on a leg press fitness case at this moment and I am trying
out hte
> paramter studies, it works fine, but my results are very jumpy. I
was
> wondering what the reason of this phenomenon could be. I attached
a
> picture in ‘files’ that shows this effect. I am only varying seat
> angle. The question is, whether these types of sharp changes are
of
> mathematical errors or biomechanical. The amount of steps is 40,
and
> which is high enough to prevent effects of too little steps. Could
you
> explain to me, why the results of the parameterstudies are so
jumpy and
> not some more smoothly divided?
>
> thank You
>
> Pamela
>