rl-collision-avoidance

This is a Pytorch implementation of the paper Towards Optimally Decentralized Multi-Robot Collision Avoidance via Deep Reinforcement Learning

Requirement

• python2.7
• ROS Kinetic
• mpi4py
• Stage --> Use this instruction
• PyTorch --> Note that this repo was modifed for non-cuda version.

How to train

Git clone this repo into catkin_ws/src and do catkin_make. All the following instructions should be executed at this repo folder (i.e. src/rl-collision-aovidance).

To train Stage1, modify the hyper-parameters in ppo_stage1.py as you like, and running the following command:

rosrun stage_ros_add_pose_and_crash stageros -g worlds/stage1.world
mpiexec -np 24 python ppo_stage1.py

To train Stage2, modify the hyper-parameters in ppo_stage2.py as you like, and running the following command:

rosrun stage_ros_add_pose_and_crash stageros -g worlds/stage2.world
mpiexec -np 44 python ppo_stage2.py

How to test

rosrun stage_ros_add_pose_and_crash stageros worlds/circle.world
mpiexec -np 50 python circle_test.py

Notice

This repo was forked from https://github.com/Acmece/rl-collision-avoidance.git. You may contact Jia Pan ([email protected]) for the paper related issues. If you find it useful and use it in your project, please consider citing:

@misc{Tianyu2018,
	author = {Tianyu Liu},
	title = {Robot Collision Avoidance via Deep Reinforcement Learning},
	year = {2018},
	publisher = {GitHub},
	journal = {GitHub repository},
	howpublished = {\url{https://github.com/Acmece/rl-collision-avoidance.git}},
	commit = {7bc682403cb9a327377481be1f110debc16babbd}
}