Structural optimization of stiffened panel structures with continuous and discrete design variables using deep reinforcement learning

Optimizing stiffened panel structures used in ships and aircrafts are becoming increasingly important for reducing material costs while maintaining or enhancing structural strength. These structures require simultaneous optimization of continuous design variables (panel thicknesses) and discrete des...

Full description

Saved in:
Bibliographic Details
Main Authors: Ryota NONAMI, Mitsuru KITAMURA
Format: Article
Language:Japanese
Published: The Japan Society of Mechanical Engineers 2025-05-01
Series:Nihon Kikai Gakkai ronbunshu
Subjects:
Online Access:https://www.jstage.jst.go.jp/article/transjsme/91/946/91_25-00020/_pdf/-char/en
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849424852695384064
author Ryota NONAMI
Mitsuru KITAMURA
author_facet Ryota NONAMI
Mitsuru KITAMURA
author_sort Ryota NONAMI
collection DOAJ
description Optimizing stiffened panel structures used in ships and aircrafts are becoming increasingly important for reducing material costs while maintaining or enhancing structural strength. These structures require simultaneous optimization of continuous design variables (panel thicknesses) and discrete design variables (stiffener cross-sectional shapes), making gradient-based methods less applicable. Consequently, heuristic approaches such as Genetic Algorithms (GA) are often employed; however, GA typically imposes a high computational burden in large-scale optimization problems. To address this, the use of deep reinforcement learning (DRL) has been investigated for large-scale combinatorial optimization. Among DRL approaches, Double Deep Q-Network (DDQN) has been reported as effective, yet its application to structural optimization involving high computational demands from Finite Element Analysis (FEA) remains limited. To overcome such an environment, in this study, an optimization flow for structural optimization is considered, and states, actions, and rewards appropriately representing design variables, constraint conditions, and objective functions are discussed. In addition, this study also proposes incorporating an elite-preservation algorithm into DDQN to reduce the computational load of structural optimization. Experimental results show that the proposed method yields designs under various load conditions that are up to 6% lighter than those obtained using GA, with a computational time reduction of approximately 81%. These findings confirm the feasibility of efficient and effective optimization for stiffened panel structures and suggest potential benefits in cost reduction and performance enhancement in future structural designs.
format Article
id doaj-art-e27f220338274c5f82592c02e444f8d1
institution Kabale University
issn 2187-9761
language Japanese
publishDate 2025-05-01
publisher The Japan Society of Mechanical Engineers
record_format Article
series Nihon Kikai Gakkai ronbunshu
spelling doaj-art-e27f220338274c5f82592c02e444f8d12025-08-20T03:29:58ZjpnThe Japan Society of Mechanical EngineersNihon Kikai Gakkai ronbunshu2187-97612025-05-019194625-0002025-0002010.1299/transjsme.25-00020transjsmeStructural optimization of stiffened panel structures with continuous and discrete design variables using deep reinforcement learningRyota NONAMI0Mitsuru KITAMURA1Department of Mechanical Engineering, National Institute of Technology (KOSEN), Kure CollegeGraduate School of Advanced Science and Engineering, Hiroshima UniversityOptimizing stiffened panel structures used in ships and aircrafts are becoming increasingly important for reducing material costs while maintaining or enhancing structural strength. These structures require simultaneous optimization of continuous design variables (panel thicknesses) and discrete design variables (stiffener cross-sectional shapes), making gradient-based methods less applicable. Consequently, heuristic approaches such as Genetic Algorithms (GA) are often employed; however, GA typically imposes a high computational burden in large-scale optimization problems. To address this, the use of deep reinforcement learning (DRL) has been investigated for large-scale combinatorial optimization. Among DRL approaches, Double Deep Q-Network (DDQN) has been reported as effective, yet its application to structural optimization involving high computational demands from Finite Element Analysis (FEA) remains limited. To overcome such an environment, in this study, an optimization flow for structural optimization is considered, and states, actions, and rewards appropriately representing design variables, constraint conditions, and objective functions are discussed. In addition, this study also proposes incorporating an elite-preservation algorithm into DDQN to reduce the computational load of structural optimization. Experimental results show that the proposed method yields designs under various load conditions that are up to 6% lighter than those obtained using GA, with a computational time reduction of approximately 81%. These findings confirm the feasibility of efficient and effective optimization for stiffened panel structures and suggest potential benefits in cost reduction and performance enhancement in future structural designs.https://www.jstage.jst.go.jp/article/transjsme/91/946/91_25-00020/_pdf/-char/endeep reinforcement learningdouble deep q-networkartifical inteligencestructural optimizationdiscrete design variables
spellingShingle Ryota NONAMI
Mitsuru KITAMURA
Structural optimization of stiffened panel structures with continuous and discrete design variables using deep reinforcement learning
Nihon Kikai Gakkai ronbunshu
deep reinforcement learning
double deep q-network
artifical inteligence
structural optimization
discrete design variables
title Structural optimization of stiffened panel structures with continuous and discrete design variables using deep reinforcement learning
title_full Structural optimization of stiffened panel structures with continuous and discrete design variables using deep reinforcement learning
title_fullStr Structural optimization of stiffened panel structures with continuous and discrete design variables using deep reinforcement learning
title_full_unstemmed Structural optimization of stiffened panel structures with continuous and discrete design variables using deep reinforcement learning
title_short Structural optimization of stiffened panel structures with continuous and discrete design variables using deep reinforcement learning
title_sort structural optimization of stiffened panel structures with continuous and discrete design variables using deep reinforcement learning
topic deep reinforcement learning
double deep q-network
artifical inteligence
structural optimization
discrete design variables
url https://www.jstage.jst.go.jp/article/transjsme/91/946/91_25-00020/_pdf/-char/en
work_keys_str_mv AT ryotanonami structuraloptimizationofstiffenedpanelstructureswithcontinuousanddiscretedesignvariablesusingdeepreinforcementlearning
AT mitsurukitamura structuraloptimizationofstiffenedpanelstructureswithcontinuousanddiscretedesignvariablesusingdeepreinforcementlearning