openai_crazyflie

This project describes how to implement Reinforcement Learning in the swarm of the Crazyflie robots system. It is built upon the openai_ros and CrazyS repositories. Note that current repo is only tested in Unbuntu 18.04 and Gazebo9.

Step 0

• Download the openai_ros related files:

    cd ~/catkin_ws/src
    git clone https://bitbucket.org/theconstructcore/openai_examples_projects.git
    git clone https://bitbucket.org/theconstructcore/openai_ros.git
    cd openai_ros
    git checkout version2
    cd ~/catkin_ws
    catkin_make
    source devel/setup.bash
  

• Download the Crazyflie related files: Please follow the steps described in CrazyS.

Step 1 Creates environments for Crazyflie robots.

• Copy the file cf_env.py to openai_ros/openai_ros/src/openai_ros/robot_envs

• Add the folder Crazyflie to openai_ros/openai_ros/src/openai_ros/task_envs

• In openai_ros/openai_ros/src/openai_ros/task_envs, add the following code at line 207.

  elif task_env == 'Crazyflie-v0':

    register(
        id=task_env,
        entry_point='openai_ros.task_envs.Crazyflie.cf_goto:CrazyflieGotoEnv',
        max_episode_steps=max_episode_steps,
    )
  
    # import our training environment
    from openai_ros.task_envs.Crazyflie import cf_goto
  

Step 2 Create the learning algorithm and main file

Use the catkin_create_pkg create the folder that contains files as in folder crazyflie_openai.

Step 3 Edit the initial position of robots

To change the number of robots and set the initial position, go to file CrazyS/rotors_gazebo/resource/spline_trajectory.yaml to change.

Step 4 First test

To run the code, firstly run roslaunch rotors_gazebo crazyflie2_swarm_hovering_example.launch. Then run roslaunch crazyflie_openai cf_start_training.launch.