I’m having some problems with my optimised model, one of the knees appears to at the wrong angle of flexion by about 180 degrees. I am using the repository v1 in AnyBody v4.0.2 and have used the GaitApplication2.exe (v1.8) with the study type POSVELACC. The marker placement model looks fine but when I run it though the gait analysis and set the initial conditions in the opt model one of the legs is completely wrong but the other leg is fine. I don’t get any errors in when I run GaitApplication2 and my kinematics will run in the opt model with the inverted knee moving.
Is it just that I haven’t got the leg set up correctly with the markers before running the GaitApplication? I have tried several times to make the leg fit the markers better but it hasn’t seemed to help. Has anyone else had this problem or got any suggestions?
There are multiple things that could lead to this. However, there is one that is more likely than the others so I think you should look into that one first:
As you may know, there are always two solutions to the problem of determining the cardan angles in a given sequence from a rotation matrix. So what you see is probably that the solver picks the opposite of the one you expect. Fortunately, it is build into GaitApplication2 that you can force it to pick the opposite solution. This is done with the keyword EULERANGLE. Please see the manual for the full explanation of this variable. In the AnyBody model, this variable is set in the DataForConfigFile.any file, where you can also change its value.
I hope that solves your problem else either write again or upload your model and we shall take a look at it.
Best regards
Michael Skipper Andersen
The AnyBody Research Project