TurtleBot3 Reinforcement Learning Project

This project demonstrates the implementation of a Q-learning algorithm enhancement to control a TurtleBot3 robot. The robot navigates different environments simulated in Gazebo, learns to reach target positions from various initial conditions, and avoids obstacles using reinforcement learning.

Table of Contents

• Prerequisites
• Installation
• Setup
• Usage
  • Empty World
  • House
  • World
• Data Availability
• Conclusion

Prerequisites

Ensure you have the following installed:

• ROS (Robot Operating System)
• Gazebo
• TurtleBot3 packages

Installation

1. Install TurtleBot3 Gazebo package:

   sudo apt-get install ros-$ROS_DISTRO-turtlebot3-gazebo

2. Set TurtleBot3 Model:

   export TURTLEBOT3_MODEL=waffle

Setup

1. Clone the repository:

   cd ~/catkin_ws/src
   git clone https://github.com/AliiRezaei/turtlebot3_rl.git

2. Build the workspace:

   cd ~/catkin_ws
   catkin_make

Usage

1. Start ROS master:

   roscore

2. Launch the simulation environments and run the learning nodes.

Empty World

1. Launch Empty World in Gazebo:

   roslaunch turtlebot3_rl turtlebot3_env_empty_world.launch

2. Run the Empty World learning node:

   rosrun turtlebot3_rl learning_empty_world

3. Wait for the robot to complete learning and observe the results.

House

1. Launch House in Gazebo:

   roslaunch turtlebot3_rl turtlebot3_env_house.launch

2. Run the House learning node:

   rosrun turtlebot3_rl learning_house

3. Wait for the robot to complete learning and observe the results.

World

1. Launch World in Gazebo:

   roslaunch turtlebot3_rl turtlebot3_env_world.launch

2. Run the World learning node:

   rosrun turtlebot3_rl learning_world

3. Wait for the robot to complete learning and observe the results.

Data Availability

Learning data is logged in the LogData directory:

• Empty World: 50,000 episodes
• House: 30,000 episodes
• World: 25,000 episodes

Conclusion

This project showcases the efficiency of a Q-learning algorithm implemented in C++ for controlling the TurtleBot3 robot. The robot demonstrates significant learning performance improvements, achieving training in approximately 30 seconds compared to 13 hours with MATLAB. For campare the MATLAB and C++ performance, run the MATLAB related script from MATLAB Scripts directory.