ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange
Ask Your Question

Does ros have some package that can provide linear and circular interpolation

asked 2017-03-08 02:22:29 -0500

yin gravatar image

updated 2017-03-08 02:23:11 -0500

hello, I am using ros to control a six-dof robot arm and can realize inverse kinematics.

Now I want to control it to move in the cartesian space, which may need linear and circular interpolation, so how could achieve this fastly as I am new about this and does ros have some package can help me do this?

Thank you for help!

edit retag flag offensive close merge delete

2 Answers

Sort by » oldest newest most voted

answered 2017-10-13 08:12:15 -0500

In the DRC we used piecewise linear cartesian motions to approximate circular motion. A helper function for this is getCircularArcPoses , which just generates poses lying on a circular arc. An example of it being called is here. The poses are then forwarded to MoveIt's cartesian planner starting here and then sent to the robot for execution.

edit flag offensive delete link more

answered 2017-10-12 21:52:45 -0500

Stephen_Z gravatar image

updated 2017-10-12 22:45:48 -0500

Linear interpolation for tool space movements is often accomplished by using the “Linear Segments with Parobolic Blends” (LSPB) method. Such a method is required to account for physical motor limitations such as instantaneous velocity changes. Refer to chapter five in this text in . Page 177 introduces Linear Segments with Parobolic Blends. Use these equations in tool space (Cartesian xyz).

this is also a good trajectory generation source

Step 1) define goal points (i.e. taught points..) Step 2) solve LSPB along the vectors connecting the goal points (solving x, y, z components independently) which gives all the needed waypoints Step 3) convert each waypoint to joint space using IK.

edit flag offensive delete link more

Question Tools



Asked: 2017-03-08 02:22:29 -0500

Seen: 947 times

Last updated: Oct 13 '17