Hypervolume-Based Multi-Objective Optimization Method Applying Deep Reinforcement Learning to the Optimization of Turbine Blade Shape

A multi-objective turbine shape optimization method based on deep reinforcement learning (DRL) is proposed. DRL-based optimization methods are useful for repeating optimization tasks that arise in applications such as the design of turbines and automotive parts. In conventional research, DRL is appl...

Full description

Saved in:
Bibliographic Details
Main Authors: Kazuo Yonekura, Ryusei Yamada, Shun Ogawa, Katsuyuki Suzuki
Format: Article
Language:English
Published: MDPI AG 2024-09-01
Series:AI
Subjects:
Online Access:https://www.mdpi.com/2673-2688/5/4/85
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850239848295694336
author Kazuo Yonekura
Ryusei Yamada
Shun Ogawa
Katsuyuki Suzuki
author_facet Kazuo Yonekura
Ryusei Yamada
Shun Ogawa
Katsuyuki Suzuki
author_sort Kazuo Yonekura
collection DOAJ
description A multi-objective turbine shape optimization method based on deep reinforcement learning (DRL) is proposed. DRL-based optimization methods are useful for repeating optimization tasks that arise in applications such as the design of turbines and automotive parts. In conventional research, DRL is applied only to single-optimization tasks. In this study, a multi-objective optimization method using improvements in hypervolume is proposed. The proposed method is applied to a benchmark problem and a turbine optimization problem. It succeeded in efficiently solving the problems, and Pareto optimal solutions are obtained.
format Article
id doaj-art-67b6e6f0533941aa970c2ac52e781616
institution OA Journals
issn 2673-2688
language English
publishDate 2024-09-01
publisher MDPI AG
record_format Article
series AI
spelling doaj-art-67b6e6f0533941aa970c2ac52e7816162025-08-20T02:01:01ZengMDPI AGAI2673-26882024-09-01541731174210.3390/ai5040085Hypervolume-Based Multi-Objective Optimization Method Applying Deep Reinforcement Learning to the Optimization of Turbine Blade ShapeKazuo Yonekura0Ryusei Yamada1Shun Ogawa2Katsuyuki Suzuki3Department of Systems Innovation, The University of Tokyo, Tokyo 113-8654, JapanDepartment of Systems Innovation, The University of Tokyo, Tokyo 113-8654, JapanDepartment of Systems Innovation, The University of Tokyo, Tokyo 113-8654, JapanDepartment of Systems Innovation, The University of Tokyo, Tokyo 113-8654, JapanA multi-objective turbine shape optimization method based on deep reinforcement learning (DRL) is proposed. DRL-based optimization methods are useful for repeating optimization tasks that arise in applications such as the design of turbines and automotive parts. In conventional research, DRL is applied only to single-optimization tasks. In this study, a multi-objective optimization method using improvements in hypervolume is proposed. The proposed method is applied to a benchmark problem and a turbine optimization problem. It succeeded in efficiently solving the problems, and Pareto optimal solutions are obtained.https://www.mdpi.com/2673-2688/5/4/85deep reinforcement learningrepeating optimization taskmulti-objective optimizationPareto solutions
spellingShingle Kazuo Yonekura
Ryusei Yamada
Shun Ogawa
Katsuyuki Suzuki
Hypervolume-Based Multi-Objective Optimization Method Applying Deep Reinforcement Learning to the Optimization of Turbine Blade Shape
AI
deep reinforcement learning
repeating optimization task
multi-objective optimization
Pareto solutions
title Hypervolume-Based Multi-Objective Optimization Method Applying Deep Reinforcement Learning to the Optimization of Turbine Blade Shape
title_full Hypervolume-Based Multi-Objective Optimization Method Applying Deep Reinforcement Learning to the Optimization of Turbine Blade Shape
title_fullStr Hypervolume-Based Multi-Objective Optimization Method Applying Deep Reinforcement Learning to the Optimization of Turbine Blade Shape
title_full_unstemmed Hypervolume-Based Multi-Objective Optimization Method Applying Deep Reinforcement Learning to the Optimization of Turbine Blade Shape
title_short Hypervolume-Based Multi-Objective Optimization Method Applying Deep Reinforcement Learning to the Optimization of Turbine Blade Shape
title_sort hypervolume based multi objective optimization method applying deep reinforcement learning to the optimization of turbine blade shape
topic deep reinforcement learning
repeating optimization task
multi-objective optimization
Pareto solutions
url https://www.mdpi.com/2673-2688/5/4/85
work_keys_str_mv AT kazuoyonekura hypervolumebasedmultiobjectiveoptimizationmethodapplyingdeepreinforcementlearningtotheoptimizationofturbinebladeshape
AT ryuseiyamada hypervolumebasedmultiobjectiveoptimizationmethodapplyingdeepreinforcementlearningtotheoptimizationofturbinebladeshape
AT shunogawa hypervolumebasedmultiobjectiveoptimizationmethodapplyingdeepreinforcementlearningtotheoptimizationofturbinebladeshape
AT katsuyukisuzuki hypervolumebasedmultiobjectiveoptimizationmethodapplyingdeepreinforcementlearningtotheoptimizationofturbinebladeshape