simtk.org

Hi,

I receive the following message at the end of the optimization, what do you think is the cause of the error and how can I fix the error?

MocoCasADiSolver did NOT succeed:
Infeasible_Problem_Detected

Hi Mohadeseh,

That error means that your problem wasn't able to meet the constraints. What problem are you trying solve?

-Nick

Thank you, Nick
Using MocoTrack, I defined the problem by following the movement of squat to stand, and the residual forces are as small as possible. I used SetStateInfo and MocoWeightState. I put contacts in instead of ground reaction forces. And it solves with the CasADi solver with a tolerance of 1e-6. The goal is to calculate the torques of the joints.
On what basis is MocoWeightState regulated? It may be because of the high coefficients that I placed for them.
Mohadeseh

Hi Mohadeseh,

Did you mean MocoWeightSet? As in, the weights applied to tracked joints in a MocoStateTrackingGoal?

If you replaced GRFS with contact spheres, how did you create an initial guess? I recommend introducing residual actuators to get an initial guess that converges before running the final problem you want to solve.

Best,
Nick

Thank you very much for your guidance.

Yes, I meant MocoWeightSet. How are these weights adjusted?

I did not take an initial guess.After a number of iterations, the solution was finished and I received the message Infeasible_Problem_Detected, but after drawing the graphs, the states tracking results were coincident. I will definitely do what you said.

Mohadeseh

Hi Mohadeseh,

The MocoWeightSet is up to you to adjust. It does not contain any decision variables to be optimized, but rather weights that are applied to the tracking terms in the cost function.

Yes, start with adding reserve and residual actuators to your model. You could be tracking the reference data well but not meeting the defect constraints.

Best,
Nick