Compound Learning Control for Formation Management of Multiple Autonomous Agents

Syafiq Fauzi, Kamarulzaman and Alsibai, Mohammad Hayyan (2016) Compound Learning Control for Formation Management of Multiple Autonomous Agents. In: International Conference on Electrical & Electronic Technology, 22-26 August 2016 , Universiti Putra Malaysia. pp. 1-6.. (Unpublished)

[img] PDF
Compound Learning Control for Formation Management of Multiple Autonomous Agents.pdf
Restricted to Repository staff only

Download (1MB) | Request a copy

Abstract

Having cooperation between multiple autonomous devices against one task is difficult due to each device having their own decision management based on self-deterministic protocol. Within the self-deterministic protocol, a formation management task should be considered along another task in order to provide cooperation and consideration between the operating autonomous devices. In this research, a compound learning control system for formation management of multiple control agents is proposed by managing coordination between multiple autonomous agents along with other tasks simultaneously in an operation. Series of simulation based on an autonomous robot was conducted to evaluate the effectiveness of learning through compound knowledge for providing consideration among achieving goals or coordination configuration against partner robot. The proposed system was able to provide consideration in coordination among operating partners in a task of achieving goal.

Item Type: Conference or Workshop Item (Speech)
Uncontrolled Keywords: learning control, multi-agent, formation management, reinforcement learning, intelligent control
Subjects: Q Science > QA Mathematics > QA76 Computer software
T Technology > T Technology (General)
Faculty/Division: Faculty of Computer System And Software Engineering
Faculty of Engineering Technology
Depositing User: Mrs. Neng Sury Sulaiman
Date Deposited: 27 Dec 2016 04:17
Last Modified: 29 Mar 2018 07:10
URI: http://umpir.ump.edu.my/id/eprint/14306
Download Statistic: View Download Statistics

Actions (login required)

View Item View Item