Skip to content

islem143/Swarm-Robotics-Self-Shape-Formation-Reinforcement-Learning

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

86 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Swarm robotics self shape formation

This repository includes my master's degree final year project in AI.

Abstarct

Swarm robotics is a fast-expanding field that includes coordinating sizable groups of robots to cooperate in order to achieve a common objective. There are multi- ple tasks and problems that swarm robots try to solve; one of them is the pattern formation problem, where the robots try to form geometric shapes. It has many applications in agriculture, search and rescue, and warehouses. One of the methods used to control the swarm is deep reinforcement learning algothims. In our project, we used the DDPG RL algorithm to control a team of robots to form some patterns with the help of a central robot that will monitor and guide the robots towards their target goals by generating the shape relative to his position. Our algorithm got an accuracy of 95% when tested on a random set of shapes when the number of robots is 3, 4. and 88% on five robots, and 72% when tested on a group of six robots.

Keywords:

Swarm robotics, Reinforcement lerarning, DDPG,Ros,Pattern foram- tion,Gazebo

Figures:

Actor Network Critic text

Training Deployment

Rewards

Shape 1

Shpae 2

Shpae 3

About

PFE MASTER

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published