Reinforcement Learning based Sequential Multi-Robot Task Allocation Considering Weight of Objects and Payload of Robots

Recently, many studies have been conducted that apply deep reinforcement learning to multi-robot task allocation. However, most of them are in the form of distributing the same number of tasks to robots, making it difficult to respond to irregular tasks. In addition, the weight of the object and the...

Full description

Saved in:
Bibliographic Details
Published in2023 23rd International Conference on Control, Automation and Systems (ICCAS) pp. 1858 - 1861
Main Authors Lee, Na-Hyun, Uhm, Tae-Young, Park, Ji-Hyun, Noh, Kyung-Seok, Kim, Hyo-Gon
Format Conference Proceeding
LanguageEnglish
Published ICROS 17.10.2023
Subjects
Automation
Control systems
Deep learning
Multi-robot task allocation
payload
Reinforcement learning
Resource management
Robots
sequential task
Task analysis
weight
Online AccessGet full text

Cover

More Information
Summary:Recently, many studies have been conducted that apply deep reinforcement learning to multi-robot task allocation. However, most of them are in the form of distributing the same number of tasks to robots, making it difficult to respond to irregular tasks. In addition, the weight of the object and the payload of the robot cannot be set differently. In this paper, we proposed reinforcement learning based sequential multi robot task allocation algorithm considering weight of objects and payload of robots. This algorithm is designed for use in environments with irregularly occurring tasks, and sequentially assigns the most suitable robot for the tasks that occur. Also, this algorithm has the advantage that it can be applied even if robots have different payloads and objects have different weights. Rainbow DQN was used for this algorithm, and the learning environment was customized using openAI gym.
ISSN:2642-3901
DOI:10.23919/ICCAS59377.2023.10316925