This is a gym env to work with the hydrone gazebo simulations, allowing the use of OpenAI Baselines and Stable Baselines deep reinforcement learning algorithms in the robot navigation training.
This project is a part of the development of some gazebo environments to apply deep-rl algorithms.
Research Gazebo environments for hydrone robot
git clone https://github.com/ricardoGrando/hydrone_deep_rl_lars
git clone https://github.com/ricardoGrando/gym_hydrone
cd gym_hydrone
pip install -e .
This is an example using the OpenAI Baselines DDPG algorithm in a hydrone environment.
.\examples\openai_baselines_ddpg.py:
import gym
import gym_hydrone
import rospy
import os
env = gym.make('hydrone_Circuit_Simple-v0', env_stage=1, observation_mode=0, continuous=True, goal_list=None)To cite this repository in publications:
@misc{gymhydrone,
author = {Grando, Ricardo},
title = {gym-hydrone},
year = {2022},
publisher = {GitHub},
journal = {GitHub repository},
howpublished = {\url{https://github.com/ricardoGrando/gym_hydrone}},
}