Staff View: Simulation of drone controller using reinforcement learning AI with hyperparameter optimization

Simulation of drone controller using reinforcement learning AI with hyperparameter optimization

Drone is one of the latest drone technologies that grows with multiple applications; one of the critical applications is for fire-fighting drones such as water hose carrying for firefighting. One of the main challenges of the drone technologies is the non-linear dynamic movement caused by a variety...

Full description

Saved in:

Bibliographic Details
Main Authors:	Abu Bakar, Mohamad Hafiz, Shamsudin, Abu Ubaidah, Abdul Rahim, Ruzairi
Format:	Conference or Workshop Item
Published:	2020
Subjects:	TK Electrical engineering. Electronics Nuclear engineering
Online Access:	http://eprints.utm.my/id/eprint/92686/ http://dx.doi.org/10.1109/ICSET51301.2020.9265381
Tags:	Add Tag No Tags, Be the first to tag this record!

id	my.utm.92686
record_format	eprints
spelling	my.utm.926862021-10-28T10:13:36Z http://eprints.utm.my/id/eprint/92686/ Simulation of drone controller using reinforcement learning AI with hyperparameter optimization Abu Bakar, Mohamad Hafiz Shamsudin, Abu Ubaidah Abdul Rahim, Ruzairi TK Electrical engineering. Electronics Nuclear engineering Drone is one of the latest drone technologies that grows with multiple applications; one of the critical applications is for fire-fighting drones such as water hose carrying for firefighting. One of the main challenges of the drone technologies is the non-linear dynamic movement caused by a variety of fire conditions. One solution is to use a nonlinear controller such as Reinforcement Learning. In this paper, Reinforcement Learning has been applied as their key control system to improve the conventional approach, which is the agent (drone) that will interact with the environment without need of the controller for the flying process. This paper is introduced an optimization method for the hyperparameter in order to achieve a better reward. In addition, we only concentrate on the learning rate (alpha) and potential reward factor discount (gamma) for optimization in this paper. From this optimization, the better performance and response from our result by using alpha = 0.1 & gamma = 0.8 with reward produced 6100 and it takes 49 seconds in the learning process. 2020 Conference or Workshop Item PeerReviewed Abu Bakar, Mohamad Hafiz and Shamsudin, Abu Ubaidah and Abdul Rahim, Ruzairi (2020) Simulation of drone controller using reinforcement learning AI with hyperparameter optimization. In: 10th IEEE International Conference on System Engineering and Technology, ICSET 2020, 9 November 2020, Shah Alam, Malaysia. http://dx.doi.org/10.1109/ICSET51301.2020.9265381
institution	Universiti Teknologi Malaysia
building	UTM Library
collection	Institutional Repository
continent	Asia
country	Malaysia
content_provider	Universiti Teknologi Malaysia
content_source	UTM Institutional Repository
url_provider	http://eprints.utm.my/
topic	TK Electrical engineering. Electronics Nuclear engineering
spellingShingle	TK Electrical engineering. Electronics Nuclear engineering Abu Bakar, Mohamad Hafiz Shamsudin, Abu Ubaidah Abdul Rahim, Ruzairi Simulation of drone controller using reinforcement learning AI with hyperparameter optimization
description	Drone is one of the latest drone technologies that grows with multiple applications; one of the critical applications is for fire-fighting drones such as water hose carrying for firefighting. One of the main challenges of the drone technologies is the non-linear dynamic movement caused by a variety of fire conditions. One solution is to use a nonlinear controller such as Reinforcement Learning. In this paper, Reinforcement Learning has been applied as their key control system to improve the conventional approach, which is the agent (drone) that will interact with the environment without need of the controller for the flying process. This paper is introduced an optimization method for the hyperparameter in order to achieve a better reward. In addition, we only concentrate on the learning rate (alpha) and potential reward factor discount (gamma) for optimization in this paper. From this optimization, the better performance and response from our result by using alpha = 0.1 & gamma = 0.8 with reward produced 6100 and it takes 49 seconds in the learning process.
format	Conference or Workshop Item
author	Abu Bakar, Mohamad Hafiz Shamsudin, Abu Ubaidah Abdul Rahim, Ruzairi
author_facet	Abu Bakar, Mohamad Hafiz Shamsudin, Abu Ubaidah Abdul Rahim, Ruzairi
author_sort	Abu Bakar, Mohamad Hafiz
title	Simulation of drone controller using reinforcement learning AI with hyperparameter optimization
title_short	Simulation of drone controller using reinforcement learning AI with hyperparameter optimization
title_full	Simulation of drone controller using reinforcement learning AI with hyperparameter optimization
title_fullStr	Simulation of drone controller using reinforcement learning AI with hyperparameter optimization
title_full_unstemmed	Simulation of drone controller using reinforcement learning AI with hyperparameter optimization
title_sort	simulation of drone controller using reinforcement learning ai with hyperparameter optimization
publishDate	2020
url	http://eprints.utm.my/id/eprint/92686/ http://dx.doi.org/10.1109/ICSET51301.2020.9265381
_version_	1715189674850385920
score	13.211869

Simulation of drone controller using reinforcement learning AI with hyperparameter optimization

Similar Items