| ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange |
Hi, I am using moveit with a UR5 robotic arm. I have joint angle goals to reach. My question is that does setting a larger goal tolerance using the function move_group.setGoalTolerance(x) helps in increasing the probability of finding a plan in a constraint environment with obstacles . Has anyone played around or experimented with this?
I have not experimented with this a lot, but the main application of the tolerances is with Pose targets. If you use bigger tolerances with JointConstraints, you can get rather unintuitive results because the tolerance applies per joint.
You might get better results in this case if your actual goal state is in collision, but part of the relaxation is not. I'm not sure it's the best idea to rely on this though, instead of setting toleranced Pose targets.
For the default planner in kinetic (RRTConnect) I do not believe it should have much of an influence on the actual trajectory planning process, because it's goal-biased either way.
Asked: 2017-09-29 01:23:59 -0500
Seen: 1,396 times
Last updated: Oct 04 '17
ROS Answers is licensed under Creative Commons Attribution 3.0 Content on this site is licensed under a Creative Commons Attribution Share Alike 3.0 license.
Also on moveit-users.