AcceleratorR/Automatic-Parking

Automatic parking with Reinforcement Learning (Q learning with epsilon greedy algorithm) in simulation. A parking environment is created in both Matplotlib and Gazebo.

Automatic Parking in simulation

This project utilized reinforcement learning algorithm (Q-learning with epsilon greedy algorithm)

Installation

Requirements:

• Python 2.7
• matplotlib
• numpy
• ROS indigo (or kinetic)

After you installed these libraries and ROS, follow the following steps:

sudo apt-get install ros-kinetic-gazebo-ros-pkgs ros-kinetic-gazebo-ros-control
sudo apt-get install ros-kinetic-ros-control ros-kinetic-ros-controllers ros-kinetic-controller-manager

mkdir ~/parking
cd ~/parking
git clone https://github.com/CTTC/Automatic-Parking.git

mkdir -p ~/catkin_ws/src
mv ackermann_model/* ~/catkin_ws/src
cd ~/catkin_ws/src
catkin_init_workspace
cd ..
catkin_make

echo "source ~/catkin_ws/devel/setup.bash" >> ~/.bashrc
source ~/.bashrc

Done! Now you can play with the ackermann car in Gazebo by:

roslaunch ackermann_vehicle_gazebo ackermann_vehicle.launch
roslaunch ackermann_drive_teleop ackermann_drive_keyop.launch

Run Demos

Demo in Matplotlib

cd ~/parking/q_learning/matplotlib_sim/demo
python agent.py

Demo in Gazebo

Start running the agent in Gazebo first:

roslaunch ackermann_vehicle_gazebo ackermann_vehicle.launch

Then run the demo

cd ~/parking/q_learning/gazebo_sim/demo
python agent.py

A presentation report is available. Demo videos are available at here and here