gloabal and local planner algorithms with learnable parameters

asked 2019-10-10 07:18:25 -0600

kiko453 gravatar image

Hi all, I'm looking for a global_planner as well as local_planner algorithms that have parameters that can be learn via machine learning technique. Does Dijkstra, A* have such parameters? Can you recommend some algorithms?

edit retag flag offensive close merge delete


All the parameters you have in any global and local planner are stored in ROS parameter server and can be accessed from any new node. You will have to think of what logic and parameters to used and how to learn. And secondly there are reconfigurable parameters that can be changed on the fly so you can modified those based on output of your model.

Choco93 gravatar image Choco93  ( 2019-10-11 05:26:40 -0600 )edit