SO and FD

Provide easy-to-use, extensible software for modeling, simulating, controlling, and analyzing the neuromusculoskeletal system.
POST REPLY
User avatar
Maurice Lang
Posts: 10
Joined: Mon Sep 25, 2017 8:42 am

SO and FD

Post by Maurice Lang » Mon Oct 30, 2017 2:24 am

Hi everyone,

I've tried to do a static optimization with a motion file in order to see muscles activations e forces,
then to verify if they should be reasonable I've done a forward dynamic simulation given as input the results (control.sto) obtained with SO, but it doesn't reproduce the original motion. Why?

Thanks

Maurice
Top
User avatar
Pierre Kibleur
Posts: 19
Joined: Tue Sep 19, 2017 2:16 am

Re: SO and FD

Post by Pierre Kibleur » Mon Oct 30, 2017 7:35 am

Hi Maurice,

I think that the SO tool should not be used on its own, since it does not consider the system's evolution in time. But for the skeleton to be in a given position, the previous muscle activations matter a lot. The SO is however called by the CMC, one time step at a time. Then, the CMC handles time integration on top of it. Examples of control-driven simulations usually use the CMC's output as input. When you will be using Forward Dynamics, make sure to double check the initial position of your system, as results will vary dramatically otherwise.

Pierre
Top
User avatar
Maurice Lang
Posts: 10
Joined: Mon Sep 25, 2017 8:42 am

Re: SO and FD

Post by Maurice Lang » Mon Oct 30, 2017 12:37 pm

Thank you for your reply!

So if I want to perform a muscle driven simulation as input for the forward dynamic I have to use the results from CMC.
But if I perform a CMC or a SO on the same problem the results are very different between them
also because SO resolves step by step while CMC integrating over time...
so how can I evaluate my results without having info from EMG, which one are more correct, from SO or CMC? Maybe if my motion is rapid it's better CMC ?

Maurice
Top
User avatar
jimmy d
Posts: 1375
Joined: Thu Oct 04, 2007 11:51 pm

Re: SO and FD

Post by jimmy d » Mon Oct 30, 2017 2:10 pm

So if I want to perform a muscle driven simulation as input for the forward dynamic I have to use the results from CMC.
yes.
But if I perform a CMC or a SO on the same problem the results are very different between them
They are different analysis, so of course, they would give different answers.
so how can I evaluate my results without having info from EMG, which one are more correct, from SO or CMC? Maybe if my motion is rapid it's better CMC ?
Give this a read; http://nmbl.stanford.edu/publications/pdf/Hicks2015.pdf
Top
POST REPLY