Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Integrate global planner with an updated controller #12

Open
wants to merge 8 commits into
base: development
Choose a base branch
from

Conversation

mcdoerr
Copy link
Collaborator

@mcdoerr mcdoerr commented Jun 15, 2021

To match the changes to the global planner, we have removed the halton points and integrated a path-following controller. The trajectory is chosen now by the cross-track-error, along-track-error, collision-cost, and heading-error to a lookahead point from the path published by the global planner in mushr_gp.

We have also added backwards rollouts to the trajectories considered in dispersion rollouts. This allows the car to follow the backwards paths sometimes generated by the global planner.

Finally, we have tuned parameters to fit the new controller, including the K number of rollouts for the dispersions.

@mcdoerr mcdoerr requested a review from schmittlema June 15, 2021 19:01
Copy link
Member

@schmittlema schmittlema left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This looks good!
Three things I noticed

  1. If you place the goal directly behind the robot but facing the opposite direction the car will not turn around. The path incorporates turning around but the car ignores it
  2. We could use more print statements. One right at startup that says "Initializing...." so that people know it is doing something. Another one for when we load a perm region from file. In case we need to delete it. This helped me with the last point below.
  3. The map is set to "default", not the map name. I think there is a warning about this. This can cause a problem if you switch maps because the permissible region will be for the previous map (default.npy). Permissible region is the region where there are no obstacles, so if you use another map's region, then you will have some funky or failing behavior.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants