Pioneer 3DX ROS Environment for Reinforcement Learning

This repository contains the ROS environment necessary for launching and controlling the Pioneer 3DX robot in a Gazebo simulation. It is specifically designed to work with a custom Gymnasium-ROS wrapper that integrates ROS and reinforcement learning (RL) for training and controlling the robot.

Table of Contents

• Overview
• Installation
• Usage
  • Launching the Environment

Overview

This environment, including the ROS map and robot configuration, is based on the DRL robot navigation repository and the work described in the publication Goal-Driven Autonomous Exploration Through Deep Reinforcement Learning.

The environment supports launching the Pioneer 3DX robot in Gazebo. It includes all necessary configurations and packages to bring up, describe, and control the robot.

Installation

Prerequisites

• ROS Noetic: Ensure that ROS Noetic is installed and sourced.
• Gazebo 11: Required for simulation.
• Python 3.8+: For using Gymnasium-based reinforcement learning environments.

Setting Up the Workspace

1. Clone this repository:

   git clone https://github.com/parisChatz/ros-rl-env.git
   cd ~/ros-rl-env/catkin_ws

2. Build the Catkin workspace:

   catkin_make_isolated

3. Source the workspace:

   source devel_isolated/setup.bash

Usage

Launching the Environment

To launch the simulation environment for the Pioneer 3DX robot:

source catkin_ws/devel_isolated/setup.bash
roslaunch multi_robot_scenario multi_robot_scenario.launch