Closed sm3304love closed 3 weeks ago
It looks like the Cartesian path goes through the collision object and if that is the case then I would expect it to fail. Since the Cartesian path is a constraint this would make the problem infeasible.
I understand. So, if I want to plan a trajectory that avoids obstacles, would it be better to refer to pick and place example?
I would start with the freepspace example.
I have confirmed through an example that to avoid obstacles, StateWaypointPoly wp1{StateWaypoint(joint_names, q)};
should be used.
The waypoint is not critical here because you should be able to use JointWaypointPoly, StateWaypointPoly, or CartesianWaypointPoly but you must configure your profile. The profile dictates what constraints and costs are added to the optimizer based on the waypoint type. A good example to review is the freespace glass upright example with uses CartesianWaypointPoly where it disable the translation constraint and only enables the rotational constraint.
Hello. To perform motion planning that avoids obstacles, I modified the basic_cartesian_example.cpp code to create a 0.1 x 0.1 x 0.1 cubic obstacle along the end effector’s path.
The function I used is as follows:
However, I encountered the following error in the terminal and motion plan failed.
When using IFOPT
When not using IFOPT