[RRA] Request for Advice on Residual Force Reduction
Posted: Fri Mar 01, 2024 2:21 am
Hello,
I am currently working on a squat motion analysis project, using the gait2354 model to work with motion capture data and force plate data.
After completing the scaling, inverse kinematics (IK), and inverse dynamics (ID) processes, I have been attempting to reduce residual forces through the Residual Reduction Algorithm (RRA).
However, I have encountered several problems during the RRA process.
Firstly, I am struggling to reduce the residual forces to zero completely, and I have also tried modifying the tasks and actuators files in this process. However, when I use the result angles (RRA_q) obtained after RRA, I experience a divergence of residual forces. (This does not happen when using the IK angles.)
I am uncertain whether this issue arises from the scaling process, IK, or the RRA itself, and I am not confident that the process I have followed so far is appropriate.
Therefore, if anyone has similar experiences or can offer advice on this issue, I would greatly appreciate your help.
Thank you.
1st RRA result is good for looking.
IK is not good but i can't improve this..
I have extensively experimented with scaling and IK, but I've struggled to find a clear direction for improvement.
I am currently working on a squat motion analysis project, using the gait2354 model to work with motion capture data and force plate data.
After completing the scaling, inverse kinematics (IK), and inverse dynamics (ID) processes, I have been attempting to reduce residual forces through the Residual Reduction Algorithm (RRA).
However, I have encountered several problems during the RRA process.
Firstly, I am struggling to reduce the residual forces to zero completely, and I have also tried modifying the tasks and actuators files in this process. However, when I use the result angles (RRA_q) obtained after RRA, I experience a divergence of residual forces. (This does not happen when using the IK angles.)
I am uncertain whether this issue arises from the scaling process, IK, or the RRA itself, and I am not confident that the process I have followed so far is appropriate.
Therefore, if anyone has similar experiences or can offer advice on this issue, I would greatly appreciate your help.
Thank you.
1st RRA result is good for looking.
- 1st RRA result
- residual_plot.png (719.29 KiB) Viewed 164 times
I have extensively experimented with scaling and IK, but I've struggled to find a clear direction for improvement.
- IK.png (193.82 KiB) Viewed 164 times