Hello,
I'm using CoppeliaSim to find an optimal robotic arm base location with respect to a specific set of target.
To do so, I have my scene with a robotic arm, the desired target dummies, and a bunch of potential base positions.
I have a script which put the arm in a base, try the IK solver from that base to all targets, and move to the next base.
It works well except that for the same base/target system, if I run the script multiple time, I may sometimes find some possible IK configs, sometimes not. I assume it is due to the IK solver which is random based, but is there a way to be more consistent ?
(For info, my model is set up exactly as in the tutorial scene "simpleManipulatorPathPlanning" for the IK part)
Have a good day,
Thibaut
Inverse Kinematic Solver
Re: Inverse Kinematic Solver
Hello Thibaut,
indeed, the function simIK.findConfigs /simIK.findConfig is randomly trying out various initial configurations, from where a desired pose can be reached via a linear motion. The thing you could do it rewrite that function in Lua, so that you can control how the random generator picks random initial configurations.
Cheers
indeed, the function simIK.findConfigs /simIK.findConfig is randomly trying out various initial configurations, from where a desired pose can be reached via a linear motion. The thing you could do it rewrite that function in Lua, so that you can control how the random generator picks random initial configurations.
Cheers