TurtleBot3

To aggregate these simulations copy the folder turtlebot3_gazebo to your turtlebot3_simulations folder to merge the files.

In the folder scripts are Python programs that we use to create the model.sdf file of the circuits. You can modify the parameters to create other circuits with diferent sizes but with the same structure.

These resouces are part of a master's thesis where we use the Python fuzzylab library to create fuzzy logic controlers with the implementation of deep reinforcement learning algorithms.

To use these resources, you first need to setup your PC following these tutorials:

1. Install Gazebo using Ubuntu packages

2. PC Setup

3. Simulation

You can apply deep reinforcement learning algorithms in the robot navigation training from OpenAI Baselines and Stable Baselines to these environments with the library gym-turtlebot3. Check the examples of the library.

Gazebo stages

turtlebot3_circuit_left_right_turns: a 5x5 stage based on the gym-gazebo GazeboCircuit2TurtlebotLidar env.

roslaunch turtlebot3_gazebo turtlebot3_circuit_left_right_turns.launch

turtlebot3_circuit_simple: a 3x3 stage.

roslaunch turtlebot3_gazebo turtlebot3_circuit_simple.launch

turtlebot3_stage_1_eavelar: a 1.8x1.8 stage based on the turtlebot3_stage_1.

roslaunch turtlebot3_gazebo turtlebot3_stage_1_eavelar.launch

This is a simple stage for simple real tests.

To cite this repository in publications:

@misc{turtlebot3environments,
  author = {Avelar, Eduardo},
  title = {Research Gazebo environments for TurtleBot3 robot},
  year = {2019},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/ITTcs/turtlebot3_simulations}},
}