corlab/rtt-core-extensions

Feeback problem for joint position controller

Closed this issue · 3 comments

smithjoshua001 commented

I've been using the lwr simulator to try and get the feedback for the joint angles, velocities, and torques, however, I noticed while using the joint position controller the feedback seems to be wrong. I've got each joint moving in a sinusoidal motion at different rates and here are the plots I get from them:
This one shows what I get when taking it directly from the LWR simulation ports.
joint_pos_simulation
This one shows what I get from the feedback on the joint position controller.
output

Both are taken using the OCL file reporter.

smithjoshua001 commented

Ah found the issue, I was looking at the wrong data, and actually it matches what the simulator provides

swrede commented

Never mind. Glad that this issues doesn't seem to be a real one... This reminds me of adding some test cases to prevent regressions. @xwavex What do you think about adding some reasonable tests?

xwavex commented

@swrede I agree, we should definitely do this!