• Search Research Projects
  • Search Researchers
  • How to Use
  1. Back to previous page

Machine learning and reinforcement learning for feature analysis of decent solution and optimization of steel structures

Research Project

Project/Area Number 18K18898
Research Category

Grant-in-Aid for Challenging Research (Exploratory)

Allocation TypeMulti-year Fund
Review Section Medium-sized Section 23:Architecture, building engineering, and related fields
Research InstitutionKyoto University

Principal Investigator

Ohsaki Makoto  京都大学, 工学研究科, 教授 (40176855)

Co-Investigator(Kenkyū-buntansha) 寒野 善博  東京大学, 大学院情報理工学系研究科, 教授 (10378812)
木村 俊明  京都大学, 工学研究科, 助教 (60816057)
Project Period (FY) 2018-06-29 – 2020-03-31
Project Status Completed (Fiscal Year 2019)
Budget Amount *help
¥5,980,000 (Direct Cost: ¥4,600,000、Indirect Cost: ¥1,380,000)
Fiscal Year 2019: ¥2,210,000 (Direct Cost: ¥1,700,000、Indirect Cost: ¥510,000)
Fiscal Year 2018: ¥3,770,000 (Direct Cost: ¥2,900,000、Indirect Cost: ¥870,000)
Keywords構造最適化 / 機械学習 / 強化学習 / 鋼構造骨組 / 建築骨組 / 構造設計 / 最適化
Outline of Final Research Achievements

A machine learning method has been developed for extracting the features related to decent mechanical performances of steel building frames. Using the proposed metod, meta-level knowledges are obtained for locations and cross-sectional properties of members, and decent designs are found with small computational cost. Furthermore, the sequential process of design update is modeled as a Markov decision process and an agent is trained using reinforment learning. As a result, an agent is developed to find designs that satisfies various design requirements. It has also been shown that the optimization method often used machine learning has better performancs for a certain type of optimum design problem than the conventional methods.

Academic Significance and Societal Importance of the Research Achievements

大規模建築鋼構造骨組の優れた力学的性能につながる部材配置や部材断面の特徴量を,機械学習の手法であるサポートベクターマシンで求める手法を提案し,その手法を用いて,静的地震荷重に対する応答を最小化する問題に対する近似最適解を,少ない計算量で得られることを示した。また,構造設計において部材断面を逐次変更するプロセスをマルコフ決定過程としてモデル化し,強化学習を用いて最適方策を求めることにより,さまざまな実際的条件を考慮した構造設計を行うエージェントを開発した。さらに,交互方向乗数法や次元削減などの機械学習で用いられる手法に基づいて,扱いにくい最適設計問題に対して優れた近似解を得る手法を提案した。

Report

(3 results)
  • 2019 Annual Research Report   Final Research Report ( PDF )
  • 2018 Research-status Report
  • Research Products

    (20 results)

All 2020 2019 2018

All Journal Article (7 results) (of which Peer Reviewed: 7 results,  Open Access: 3 results) Presentation (13 results) (of which Int'l Joint Research: 4 results)

  • [Journal Article] Dimensionality reduction enhances data-driven reliability-based design optimize2020

    • Author(s)
      Y. Kanno
    • Journal Title

      Journal of Advanced Mechanical Design, Systems, and Manufacturing

      Volume: 14 Issue: 1 Pages: JAMDSM0008-JAMDSM0008

    • DOI

      10.1299/jamdsm.2020jamdsm0008

    • NAID

      130007792197

    • ISSN
      1881-3054
    • Related Report
      2019 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] Reinforcement learning for optimum design of a plane frame under static loads2020

    • Author(s)
      Kazuki Hayashi and Makoto Ohsaki
    • Journal Title

      Engineering with Computers

      Volume: - Issue: 3 Pages: 1999-2011

    • DOI

      10.1007/s00366-019-00926-7

    • Related Report
      2019 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Reinforcement learning and graph embedding for binary truss topology optimization under stress and displacement constraints.2020

    • Author(s)
      K. Hayashi and M. Ohsaki
    • Journal Title

      Frontiers in Built Environment, Specialty Section: Computational Methods in Structural Engineering

      Volume: 6

    • DOI

      10.3389/fbuil.2020.00059

    • NAID

      120006940296

    • Related Report
      2019 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] Mixed-integer programming formulation of a data-driven solver in computational elasticity.2019

    • Author(s)
      Y. Kanno
    • Journal Title

      Optimization Letters

      Volume: 13 Issue: 7 Pages: 1505-1514

    • DOI

      10.1007/s11590-019-01409-w

    • Related Report
      2019 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Simple heuristic for data-driven computational elasticity with material data involving noise and outliers: a local robust regression approach2018

    • Author(s)
      Y. Kanno
    • Journal Title

      Japan Journal of Industrial and Applied Mathematics

      Volume: 35 Issue: 3 Pages: 1085-1101

    • DOI

      10.1007/s13160-018-0323-y

    • NAID

      210000188236

    • Related Report
      2018 Research-status Report
    • Peer Reviewed
  • [Journal Article] Data-driven computing in elasticity via kernel regression2018

    • Author(s)
      Y. Kanno
    • Journal Title

      Theoretical and Applied Mechanics Letters

      Volume: 8 Issue: 6 Pages: 361-365

    • DOI

      10.1016/j.taml.2018.06.004

    • Related Report
      2018 Research-status Report
    • Peer Reviewed / Open Access
  • [Journal Article] Alternating Direction Method of Multipliers as Simple Heuristic for Topology Optimization of a Truss With Uniformed Member Cross Sections2018

    • Author(s)
      Kanno Yoshihiro
    • Journal Title

      Journal of Mechanical Design

      Volume: 141 Issue: 1 Pages: 011403-011403

    • DOI

      10.1115/1.4041174

    • Related Report
      2018 Research-status Report
    • Peer Reviewed
  • [Presentation] Minimum-volume design of steel frames using reinforcement learning, Proc.2020

    • Author(s)
      K. Hayashi and M. Ohsaki
    • Organizer
      14th World Congress in Computational Mechanics (WCCM-ECCOMAS 2020)
    • Related Report
      2019 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Machine learning for approximate optimal placement of braces of plane steel frames under static loads.2020

    • Author(s)
      K. Sakaguchi and M. Ohsaki
    • Organizer
      Asian Congress of Structural and Multidisciplinary Otimization (ACSMO 2020)
    • Related Report
      2019 Annual Research Report
    • Int'l Joint Research
  • [Presentation] パレートランクの機械学習を用いた鋼構造骨組のブレース配置の分析と多目的最適化2020

    • Author(s)
      岩越雄一, 大崎 純, 阪口一真
    • Organizer
      日本建築学会近畿支部研究報告集
    • Related Report
      2019 Annual Research Report
  • [Presentation] パレートランクの機械学習を用いた鋼構造骨組の設計の多目的最適化2020

    • Author(s)
      岩越雄一, 大崎 純, 阪口一真
    • Organizer
      日本建築学会大会学術講演梗概集
    • Related Report
      2019 Annual Research Report
  • [Presentation] グラフ埋め込みと強化学習による鋼構造平面骨組の断面設計2020

    • Author(s)
      林 和希,大崎 純
    • Organizer
      日本建築学会大会学術講演梗概集
    • Related Report
      2019 Annual Research Report
  • [Presentation] Deep-Q network for truss topology optimization with stress constraints Proc.2019

    • Author(s)
      K. Hayashi and M. Ohsaki
    • Organizer
      IASS Symposium 2019, Barcelona, Spain, Int. Assoc.
    • Related Report
      2019 Annual Research Report
    • Int'l Joint Research
  • [Presentation] 機械学習による小規模ブレース付骨組の特徴分析とそれに基づく大規模骨組の最適化2019

    • Author(s)
      阪口 一真, 大崎 純, 木村俊明
    • Organizer
      第42回情報・システム・利用・技術シンポジウム
    • Related Report
      2019 Annual Research Report
  • [Presentation] 機械学習を用いた大規模鋼構造骨組のブレース配置の性能予測2019

    • Author(s)
      阪口一真, 大崎 純, 木村俊明
    • Organizer
      日本建築学会近畿支部研究報告集
    • Related Report
      2018 Research-status Report
  • [Presentation] 機械学習を用いた鋼構造骨組の最適ブレース配置の特徴抽出2019

    • Author(s)
      阪口一真, 大崎 純, 木村俊明
    • Organizer
      日本建築学会大会学術講演梗概集(北陸)
    • Related Report
      2018 Research-status Report
  • [Presentation] Machine learning for selection of approximate optimal placement of braces of plane frames under static loads2019

    • Author(s)
      M. Ohsaki, T. Kimura and K. Sakaguchi
    • Organizer
      Proc. 13th World Congress of Structural and Multidisciplinary Optimization (WCSMO13)
    • Related Report
      2018 Research-status Report
    • Int'l Joint Research
  • [Presentation] 機械学習を用いた鋼構造骨組のブレース配置の性能予測と組合せ最適化2018

    • Author(s)
      田村拓也,大崎 純,木村俊明,高木次郎
    • Organizer
      日本建築学会大会学術講演梗概集(東北)
    • Related Report
      2018 Research-status Report
  • [Presentation] 動的計画法を用いたラーメン構造の形状最適化2018

    • Author(s)
      林 和希,大崎 純
    • Organizer
      第41回情報・システム・利用・技術シンポジウム,日本建築学会・情報システム技術委員会
    • Related Report
      2018 Research-status Report
  • [Presentation] 機械学習による応答予測を用いた鋼構造ブレース補強骨組の付加応力最小化2018

    • Author(s)
      木村俊明,大崎 純,田村拓也,高木次郎
    • Organizer
      第41回情報・システム・利用・技術シンポジウム,日本建築学会・情報システム技術委員会
    • Related Report
      2018 Research-status Report

URL: 

Published: 2018-07-25   Modified: 2021-02-19  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi