simtk.org

Hello all,
I have a little question, I think it is not difficult to answer for you:
I recently worked with the model gait2392. But now I want to reduce the amount of muscles used in this model and I just want to have four muscles for each leg (knee extension/bending). It was not difficult for me to delete these muscles from the model-file. But I'm not sure, which other settings I need to change to get the model to work. By now, RRA stops due to an overconstrained system. In the Control-Constraints file I only left FX, FY, FZ, MX, MY, MZ and knee_angle_l, knee_angle_r. Is this right? What do I need to change else?

Greetings from Germany,
Daniel

Hi,
you have an overconstrained system because the number of degrees of freedom is greater than the number of muscles. So the minimum num of muscles must be the same of the dofs to actuate the model.
Cheers,
Giordano.

Hey Giordano,
thats clear. But I don't know how to reduce the number of dofs in the model-file. Which tags do I have to modify?
Greets,
Daniel

You could modify the model file, but it would be much easier to simply "lock" the dofs (i.e., coordinates) that you no longer need. You can lock coordinates in the Coordinates window.

I see. I have the following problem: I want to reduce the model to simulate a squat jump. For this, I just want to use four muscles each leg without all the other muscles. I deleted the muscles from the model file. If I run RRA, everything is fine. But running CMC, the model performs absolutely senseless movements. I don't know, what the reason might be. I locked the joints I don't need and just left knee angles, hip flexion and t_y unlocked.
Can you imagine, what the reason for my problem might be?

I experimented a long time with the weight settings in the task file. By adjusting the settings I'm getting nearer to the original movement. What I don't understand: Why does the algorithm alter the movement? Why doesn't it "just" take the movement to compute the muscle excitations?
I would be glad, if anyone can explain this to me...