search

AtsushiSakai / PythonRobotics

Python sample codes for robotics algorithms.
https://atsushisakai.github.io/PythonRobotics/
Other
23.27k stars 6.54k forks source link

Dubin's goal doesn't end where expected #276

Closed tsturzl closed 4 years ago

tsturzl commented 4 years ago

Describe the bug DubinsPath goal has a margin of error, seems to be off by 0.2m

Expected behavior DubinsPath last point is equal to the goal

Screenshots If applicable, add screenshots to help explain your problem.

Desktop (please complete the following information):

tsturzl commented 4 years ago

I've ported the dubins path implementation to another language and I'm seeing the same issue. I'm wondering if there's a reason for this, and if this is expected behavior. I've looked at RRT Dubins and I noticed you're just moving the RRT nodes to where the dubins path ends. If some explanation can be given as to why this is the case I'd be happy to try and submit a PR if that is helpful.

tsturzl commented 4 years ago

this is what I'm seeing . I'm just going from point a to point b and then looping back from point b to point a, I'd expect this to make a closed loop, but the end of each path is not exactly located on the goal.

AtsushiSakai commented 4 years ago

@tsturzl Hi. Thank you for your report. I believe that is because of accumulated error of generate_course function: https://github.com/AtsushiSakai/PythonRobotics/blob/123c4f777ccfc49a0a11a8b99c9a63013d46b9a7/PathPlanning/DubinsPath/dubins_path_planning.py#L214

AtsushiSakai commented 4 years ago

I will fix this issue on #277

AtsushiSakai commented 4 years ago

@tsturzl Hi. I believe I fixed this issue. If you find this bug again, please reopen it. Thanks.

tsturzl commented 4 years ago

I'll take a look at your changes. A lot of your work has been really helpful, and I've been using it as a reference for a few of my own implementations. I appreciate the work you do.