Project - Differential Drive Robot with A* Path Planner

• This Project Implements A* Path Planning Algorithm for a differential drive robot.
• This Project is also simulated on Turtlebot 2 using ROS Gazebo
• The Project is implemented on real Turtlebot3 Robot at Robot Realization Laboratoy(RRL), University of Maryland, College Park

Map of the environment

The Following Map is used to plan the path

Map of RRL Lab

Dependencies

following packages have to be imported for running this file

• python 3.5 or above
• Gazebo 7 or above
• ROS Kinetic
• numpy
• math
• queue
• argparse
• matplotlib.pyplot

Execution

The code is divided into four parts

main.py

obstacle_space_offset.py

action_set_offset.py

publisher.py

• The main.py imports the action_set_offset.py and obstacle_space_offset.py while it is being executed. The input for executing this needs to be given as following

python main.py [start_position] [goal_position]

• The sample imput for running main.py are given as below

python main.py 1,2 10,10

python main.py 0,0 15,30 --> used for generating the video (this is generated for low rpm's, rpm1 = 2 and rpm2 = 3)

• It's good to take rpm1 = 5 and rpm2 = 4 for relatively efficient performance of the algorithm

• After this program is executed, it will generate the converted_velocities.txt file which will be supplied to the ROS Gazebo.

• Now open a new terminal to initialize the roscore

• A ROS package has to be created. The package should have the launh file and the rrl_map. Instructions on how to create a package and write launch files can be found here : http://wiki.ros.org/ROS/Tutorials/CreatingPackage

• In my case the name of the package that I have created is planning.

• Once, the package is successfully created, open a new terminal, launch the ROS gazebo using the following launch commands to load the rrl_map as following

ROBOT_INITIAL_POSE="-x [x coordinated of start position] -y [y coordinated of the start position]" roslaunch planning sru.launch This will spawn the turtlebot at the start position.

note: the gazebo takes the spawn coordinated in meters, hence we need to give start_position divided by 10 which is my resolution

The sample command to launch the environment is ROBOT_INITIAL_POSE="-x 0.1 -y 0.2" roslaunch planning sru.launch

• Open a new terminal to run the publisher.py file as following rosrun [package-name] publisher.py

The sample command to start the node.rosrun planning publisher.py

Results

The results are located in the output/

See how it works

TurtleBot 2 Simulated in Gazebo