Hi everyone,
I'm actually trying to implement a potential field algorithm for path planning in pybullet.
I have the model of a multibody imported from an .urdf file and an obstacle. I have already
computed the world position of closest points with respect to the obstacle for each body
of the multibody (updated in real-time).
Now, I would like to get the jacobian for each of those points, this is how I think I should do:
- Substract the world vector of the CoM with the world vector of the point
- Transform this vector into the local frame (because localPosition should be in local coordinates)
How can I transform this vector ? Also, how to manage my fixed base when I should send
the list of joint positions ? Because actually, my first jacobian is full of 0.
Thank you very much,
Youssef.