[ros-users] trajectory error in simulation / controller parameters

Gil Jones gjones at willowgarage.com
Fri Feb 11 18:36:01 UTC 2011

Hi Jannik,

The path you are seeing is indeed the path that comes from the planner.  But
there's an extra step that move_arm goes through of filtering the trajectory
using a spline_shortcutting method that tries to find ways to make the path
shorter while also assigning reasonable velocities when it can.  This can
change the trajectory quite a bit.  If you want to see the path that
move_arm will actually be sending to be executed you'll need to change the
code a bit.  Overlay arm_navigation.  In whatever version of the code you
are running find the line:

visualizePlan(current_trajectory_);  //move_arm_simple_action.cpp:1361 in

Comment this line out, and then copy it an add it after the line:

current_trajectory_ = filtered_trajectory; //move_arm_simple_action.cpp:1422
in unstable
visualizePlan(current_trajectory_); // copied from 1361

The trajectory you see in rviz will now be the actual one getting passed to
the controller.  I expect it will look much more like the executed

Let us know if you have other questions.


E. Gil Jones (gjones at willowgarage.com)
Research Engineer
Willow Garage, Inc.
68 Willow Road
Menlo Park, CA 94025
650.475.9772 <+16504759772>

On Fri, Feb 11, 2011 at 10:18 AM, Jannik Kett <jannik.kett at googlemail.com>wrote:

> Hello everyone,
> I'm a student at University of Stuttgart and work at IPA Fraunhofer with
> the Care-O-bot.
> We have here some problems with the planned trajectory in simulation. The
> actual executed trajectory of the robot arm in rviz differs from the planned
> one (display_path (motion_planning_msgs/DisplayTrajectory)). We guess that
> it comes from the dynamic model and the chosen controller parameters in the
> arm_controller_cob3-1.yaml file
> (cob_simulation/cob_controller_configuration_gazebo/ros/controller).
> So we're looking for ideas, hints and rules how to choose these parameters
> or explanations where the trajectory error could come from.
> Thanks in advance for your help!
> -Jannik
> _______________________________________________
> ros-users mailing list
> ros-users at code.ros.org
> https://code.ros.org/mailman/listinfo/ros-users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ros.org/pipermail/ros-users/attachments/20110211/3aacab7e/attachment-0003.html>

More information about the ros-users mailing list