0
$\begingroup$

I am studying motion planning, and I am focusing on artificial potential fields. I have seen that it consists in having a repulsive field for obstacles, and an attractive field toward the goal.

But what I don't understand is if this motion planning method creates a path which can be tracked, or it is just a matter of potential fields, and so the robot moves accordingly to them and not on a path.

Can somebody help me understand this?

$\endgroup$
0
$\begingroup$

It is a bit of both. The repulsive potentials help create a way to represent an environment that is conducive to planning. In the area of repulsive potentials, a simple gradient descent from a point will take you to the minimum, i.e., some collision free point.

At the same time, the attractive potential increases as you move away from the goal, so attempting to minimize that function will eventually lead you towards the goal. For any given goal state, minimizing both these potentials will result in a path that can then be tracked using a controller of your choice. For a more intuitive understanding, you might want to look at a sample implementation:

Potential fields from the PythonRobotics repository

Howie Choset's slides

| improve this answer | |
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.