CMC failed

[Daniele Tregnaghi]

Hi all, I' ve got a problem during the execution of CMC. The following message displays:

SimTK Exception thrown at InteriorPointOptimizer.cpp:249:
Optimizer failed: Ipopt: Infeasible problem detected (status 2)
OPTIMIZATION FAILED...

rdCMC.computeControls: WARN- The optimizer could not find a solution at time = 0.010000.
If using the fast target, try using the slow target.
May the problem be I have different frequencies between markers and grf? Or grf frequencies are too high? (more than 100 T=0.01)
Thanks

[Daniele Tregnaghi]

Sorry, I solve this problem.
Another question: to compute CMC for a single gait cycle, does the grf must be applied to both ankle (calcn_r calcn_l)or not? And how does CMC tool work during the double swing phase (run cycle) when no grf are applied?
Thanks

[Ayman Habib]

Daniele,

You need to apply the forces measured to proper segments (calcn_r or calcn_l). In single limb support, one of the grf's will be 0. You can have two forces with value of 0 for the flight phase of running.

Hope this helps,
-Ayman

[Daniele Tregnaghi]

Thank you.
Now, I'm analyzing a squat motion, I've got two grf applied to calcn_r and calcn_l from t=0. But CMC always fails although I change time range to process and the follow message displays:

SimTK Exception thrown at InteriorPointOptimizer.cpp:249:
Optimizer failed: Ipopt: Maximum iterations exceeded (status -1)
OPTIMIZATION FAILED...

I do not understand what should I do to solve the problem.

Any suggestion?
Daniele

[James Chalfant]

I am getting the same error as described in the first post:

SimTK Exception thrown at InteriorPointOptimizer.cpp:249:
Optimizer failed: Ipopt: Infeasible problem detected (status 2)
OPTIMIZATION FAILED...

rdCMC.computeControls: WARN- The optimizer could not find a solution at time = 0.010000.
If using the fast target, try using the slow target.

Could anyone tell me how to fix the problem?

[Yi Wang]

Hi Daniele, would you mind giving some suggestion how to fix the infeasible problem?

I found that the GRF capture rate in Gait2354 it 600Hz, which is the same with my data, so I don't think the GRF data is that high to comply with the Kinematics in CMC step.

Any clues or suggestions would help, thanks in advance.

[aurelien liger]

Hello,

I've got the same problem. Have you managed to solve it ?

I've tried a lot oh things, especially according to https://simtk.org/forum/message.php?msg_id=3427 but still not ok.

If anyone can help me?

Thanks,

[Yi Wang]

James, Daniele, Aurelien

Sorry to reply this old thread, but I can't help to ask about your RRA step. Whether have you noticed that the RRA step have changed your model's kinematics? My RRA runs perfect except that from the HEEL STRIKE point the gait2354 keep going down at the same time of moving forward.

So after the RRA in the result file xxx_q.sto which located in the RRA result directory, this new kinematics file make my model move down with moving forwards.

Have you gotten any ideas about what's wrong with my RRA? Or I could turn to previous steps such like SCALE, IK and ID, ID is pretty when I observed the curve generated from the ID results.

Thanks in advance. I have stuck in this confusion over weeks.

Yi.

PS:the forum could not be searched by keyword is not of convenience indeed.

[Ruoli Wang]

Hi Yi,

I didn't know if you already fixed the problem with walking down. I had the same problem before and fixed it by increase the tracking weight of pevlis_ty. And another thing also help me to solve the problem. Since we only had two force plates in the lab, so actually only part of the gait phase can be simulated e.g start with singe limb stance phase. When I corrected the time range, the RRA looks much better.

[Max Rusch]

Hi Daniele,

How do you solve the previous problem?

I had the same problem using the upper limb model during the static optimization tool.

Regards, Max