Computational Technology Resources - CCC

X. Zhuang¹, W. Zhang^1,2, X. Guo^1,2 and S.-K. Youn^3,1

¹State Key Laboratory of Structural Analysis for Industrial Equipment, Department of Engineering Mechanics, International Research Center for Computational Mechanics, Dalian University of Technology, China
², Ningbo Institute of Dalian University of Technology, Jiangbei District, Ningbo, China
³Department of Mechanical Engineering, KAIST, Daejeon, Republic of Korea

X. Zhuang, W. Zhang, X. Guo, S.-K. Youn, "Machine Learning-Powered Geometry-Aware Filter: A Novel Human-Informed Approach for Advanced Topology Optimization", in P. Iványi, J. Kruis, B.H.V. Topping, (Editors), "Proceedings of the Seventh International Conference on Artificial Intelligence, Soft Computing, Machine Learning and Optimization in Engineering", Civil-Comp Press, Edinburgh, UK, Online volume: CCC 11, Paper 4.2, 2025,

Keywords: topology optimization, filtering techniques, geometry filter, geometric features, human-Informed, machine learning.

Abstract

This paper proposes a machine learning-powered Geometry Filter to enhance the design capabilities of SIMP-based topology optimization. By reconstructing the paradigm of traditional density filtering functions, the method establishes a filtering format with geometric feature modulation, enabling effective interaction between human design intent and the topology optimization process. To achieve this, machine learning is employed to construct multimodal geometric feature matching metrics, transforming explicit geometric elements, abstract stylistic features, and intuitive conceptual designs into mathematically embeddable representations within the optimization workflow. Through a dynamic mapping mechanism between geometric features and density fields, the filter evolves beyond a mere numerical tool for stabilizing optimization instabilities, becoming an active geometric feature modulation component in topology optimization. Since human design intent is directly embedded in the filter, the need for additional complex feature constraints is eliminated, significantly reducing optimization complexity. Numerical examples demonstrate the method’s flexibility in generating structures with diverse geometric features, effectively facilitating human-machine interaction between design intent and structural mechanical performance.

download the full-text of this paper (PDF, 22 pages, 1319 Kb)

go to the previous paper
go to the next paper
return to the table of contents
return to the volume description

	Computational & Technology Resources an online resource for computational, engineering & technology publications
	not logged in - login
Front Page Browse CCC CCP CSETS CTR IJRT Other Authors Search Purchase Guide FAQ Contact us	Civil-Comp Conferences ISSN 2753-3239 CCC: 11 PROCEEDINGS OF THE SEVENTH INTERNATIONAL CONFERENCE ON ARTIFICIAL INTELLIGENCE, SOFT COMPUTING, MACHINE LEARNING AND OPTIMIZATION IN ENGINEERING Edited by: P. Iványi, J. Kruis and B.H.V. Topping Paper 4.2 Machine Learning-Powered Geometry-Aware Filter: A Novel Human-Informed Approach for Advanced Topology Optimization X. Zhuang¹, W. Zhang^1,2, X. Guo^1,2 and S.-K. Youn^3,1 ¹State Key Laboratory of Structural Analysis for Industrial Equipment, Department of Engineering Mechanics, International Research Center for Computational Mechanics, Dalian University of Technology, China ², Ningbo Institute of Dalian University of Technology, Jiangbei District, Ningbo, China ³Department of Mechanical Engineering, KAIST, Daejeon, Republic of Korea Full Bibliographic Reference for this paper X. Zhuang, W. Zhang, X. Guo, S.-K. Youn, "Machine Learning-Powered Geometry-Aware Filter: A Novel Human-Informed Approach for Advanced Topology Optimization", in P. Iványi, J. Kruis, B.H.V. Topping, (Editors), "Proceedings of the Seventh International Conference on Artificial Intelligence, Soft Computing, Machine Learning and Optimization in Engineering", Civil-Comp Press, Edinburgh, UK, Online volume: CCC 11, Paper 4.2, 2025, Keywords: topology optimization, filtering techniques, geometry filter, geometric features, human-Informed, machine learning. Abstract This paper proposes a machine learning-powered Geometry Filter to enhance the design capabilities of SIMP-based topology optimization. By reconstructing the paradigm of traditional density filtering functions, the method establishes a filtering format with geometric feature modulation, enabling effective interaction between human design intent and the topology optimization process. To achieve this, machine learning is employed to construct multimodal geometric feature matching metrics, transforming explicit geometric elements, abstract stylistic features, and intuitive conceptual designs into mathematically embeddable representations within the optimization workflow. Through a dynamic mapping mechanism between geometric features and density fields, the filter evolves beyond a mere numerical tool for stabilizing optimization instabilities, becoming an active geometric feature modulation component in topology optimization. Since human design intent is directly embedded in the filter, the need for additional complex feature constraints is eliminated, significantly reducing optimization complexity. Numerical examples demonstrate the method’s flexibility in generating structures with diverse geometric features, effectively facilitating human-machine interaction between design intent and structural mechanical performance. download the full-text of this paper (PDF, 22 pages, 1319 Kb) go to the previous paper go to the next paper return to the table of contents return to the volume description
Back to top	©Civil-Comp Limited 2023 - terms & conditions