Home | Tutorials | Wiki | Issues
Ask Your Question
0

Simulating an impedance controller in gazebo

asked 2012-11-09 14:09:46 -0500

Teuntjez gravatar image

Hi,

I'm trying to simulate an impedance controller for a manipulator in Gazebo. Therefore I need Cartesian force measurement on the gripper, I'm doing this using the bumper plugin. I modeled an object which I should follow with constant force (=/0) with a specific trajectory. When in static contact with the object I get a good force measurement but as soon as I input the trajectory the grippoint is pulled out of the object and the force goes to zero. When executing the same trajectory without the object the reference trajectory is executed nicely. I'm simulating this in the empty.world file, using the following settings: <physics:ode> <stepTime>0.001</stepTime> <gravity>0 0 -9.8</gravity> <cfm>0.0000000001</cfm> <erp>0.2</erp> <quickStep>true</quickStep> <quickStepIters>280</quickStepIters> <quickStepW>1.3</quickStepW> <contactMaxCorrectingVel>100.0</contactMaxCorrectingVel> <contactSurfaceLayer>0.001</contactSurfaceLayer> <updateRate>100000000000000000000</updateRate> </physics:ode>

The objects contact stiffness is equal to 100 N/m. So the question is why doesn't Gazebo allow me to stay "inside" the object but wants to pull out to reach zero contanct force. Thanks in advance

edit retag flag offensive close merge delete

Comments

How are you applying the trajectory? Are you simply prescribing the position or are you using a controller (/prescribing a joint force or acceleration).

Erik Stoltenborg gravatar imageErik Stoltenborg ( 2012-11-13 03:08:53 -0500 )edit

If you are just pescribing position i think this might be the cause of your error. The gripper then `teleports' into the object, hence gets pushed out. Instead of gradually pushing in.(contact is basically simulated as a spring damper)

Erik Stoltenborg gravatar imageErik Stoltenborg ( 2012-11-14 10:42:46 -0500 )edit

Hey, I'm prescribing a position in the gripper end point Cartesian frame. I already fixed the problem by making a model of a spring, using two masses connected with a prismatic joint with limits ( =0) which gives a constraint violation error linear dependent on simulation parameters.

Teuntjez gravatar imageTeuntjez ( 2012-11-19 02:38:22 -0500 )edit

You should post your answer on here, because others might profit from this knowledge. For what it is worth, i think the problem would not occur when you describe the motion in a `continuous' way.

Erik Stoltenborg gravatar imageErik Stoltenborg ( 2012-11-21 09:04:17 -0500 )edit

1 Answer

Sort by ยป oldest newest most voted
0

answered 2013-01-11 18:18:51 -0500

scpeters gravatar image

From comment above:

Hey, I'm prescribing a position in the gripper end point Cartesian frame. I already fixed the problem by making a model of a spring, using two masses connected with a prismatic joint with limits ( =0) which gives a constraint violation error linear dependent on simulation parameters.

Teuntjez (Nov 19 '12)

edit flag offensive delete link more
Login/Signup to Answer

Question Tools

Stats

Asked: 2012-11-09 14:09:46 -0500

Seen: 395 times

Last updated: Jan 11 '13