search

ZJU-FAST-Lab / ego-planner-swarm

An efficient single/multi-agent trajectory planner for multicopters.
GNU General Public License v3.0
1.13k stars 222 forks source link

Ego planner map size constraints #23

Closed shubham-shahh closed 2 years ago

shubham-shahh commented 2 years ago

Hi, @bigsuperZZZX The planner is constrained by the map size that is initially given. here the size of the buffer becomes large which limits the maximum size.

are there any approaches to make the planner free from the map size constraints?

thanks

bigsuperZZZX commented 2 years ago

Using a 3D-circular buffer can solve this. You can code by yourselves or wait for our brand new planner along with a ring map which will be released in about 2 months.

shubham-shahh commented 2 years ago

Using a 3D-circular buffer can solve this. You can code by yourselves or wait for our brand new planner along with a ring map which will be released in about 2 months.

okay, thanks for the update

shubham-shahh commented 2 years ago

Hi @bigsuperZZZX, is it necessary for this vector to store double type data? is this much amount of precision required?

I found the range of values stored in the md_.occupancy_buffer_ variable ranges from -2.00243 to 2.197, can you please explain me what does this values translate to? I thought it must be storing probability ranging from 0-1

thanks

bigsuperZZZX commented 2 years ago

It is log-odds representation not the original probability value.

And using double is not necessary indeed.

shubham-shahh commented 2 years ago

thanks for the information.

shubham-shahh commented 2 years ago

Using a 3D-circular buffer can solve this. You can code by yourselves or wait for our brand new planner along with a ring map which will be released in about 2 months.

Hi @bigsuperZZZX, I am curious if there are any updates to the latest paper

Thanks

bigsuperZZZX commented 2 years ago

You would like to refer to this repo