...e joint impedance control strategy. That is, the robot behaves as a spring with stiffness K which is fixed for now.
This commit differentiates between joint positions controlled by effort (requires PID tunning) and controlled by means of joint stiffness (make the most of the impedance capabilities of the arm).
The stiffness value is set in the initialization of the joint_stiffness_command member here. I'm already using this in the multi-robot-test branch with MoveIt, and it is much more smooth.
For the moment, the stiffness value is hardcoded, but I plan to create a position controller for that, whence the _command, such that one can eventually control stiffness trajectories, and synchronize them with position trajectories.
I'm creating the pull request just to announce it, I will merge it in a couple of days.
carlosjoserg
commented
9 years ago
...e joint impedance control strategy. That is, the robot behaves as a spring with stiffness K which is fixed for now.
This commit differentiates between joint positions controlled by effort (requires PID tunning) and controlled by means of joint stiffness (make the most of the impedance capabilities of the arm).
The stiffness value is set in the initialization of the
joint_stiffness_commandmember here. I'm already using this in themulti-robot-testbranch with MoveIt, and it is much more smooth.For the moment, the stiffness value is hardcoded, but I plan to create a position controller for that, whence the
_command, such that one can eventually control stiffness trajectories, and synchronize them with position trajectories.I'm creating the pull request just to announce it, I will merge it in a couple of days.