簡易檢索 / 詳目顯示

回結果列表
研究生： 曾柏諺
Tseng, Bor-Yann
論文名稱： 應用深度學習與強化學習於結構材料的設計
Application of deep learning and reinforcement learning in structural materials by design
指導教授： 游濟華
Yu, Chi-Hua
學位類別： 博士
Doctor
系所名稱： 工學院 - 工程科學系
Department of Engineering Science
論文出版年： 2025
畢業學年度： 113
語文別： 英文
論文頁數： 160
中文關鍵詞： 人工智慧 、深度學習 、強化學習 、結構材料 、反向設計
外文關鍵詞： artificial intelligence, deep learning, reinforcement learning, structural materials, inverse design
相關次數： 點閱：153  下載：0 
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報
    • 摘要 i Abstract ii 致謝 iii Table of Contents iv Table of Tables viii Table of Figures ix Chapter 1 Introduction 1 1.1 Background 1 1.2 Motivation 3 1.3 Objectives 6 1.4 Organization 8 Chapter 2 Literature Review 9 2.1 Q-Learning and Deep Q-Learning 9 2.2 Deep Deterministic Policy Gradient 12 2.3 ResNet 13 2.4 Unet 15 2.5 Generative Adversarial Network (GAN) 17 2.6 Transformer 19 2.7 Consistency Models 21 2.8 Composites Design 22 2.9 Inverse Design 24 Chapter 3 Composites Design using Reinforcement Learning 26 3.1 Application of Reinforcement Learning 26 3.2 Materials and Methods 28 3.2.1 Composites design using progressive reinforcement learning 28 3.2.2 Impact resistant component design using reinforcement learning 31 3.2.3 Local structural design using reinforcement learning 33 3.3 Results and Discussion 35 3.3.1 Results of global design of composite materials 35 3.3.2 Results of impact resistant component optimize 38 3.3.3 Results of nacre like structure optimize 40 3.4 Chapter Summary 42 Chapter 4 Predicting and Designing Dendritic Structures Using Deep Learning and Reinforcement Learning 44 4.1 Prediction and Control of Dendritic Structures Growth 44 4.2 Materials and Methods 45 4.2.1 Phase field method 46 4.2.2 Dataset of dendritic structures 48 4.2.3 Predict the architecture of dendritic structures using ResUnet 50 4.2.4 Porosity control using reinforcement learning 51 4.3 Results 52 4.3.1 Results of single nucleus generation 53 4.3.2 Results of multi-nucleus generation 55 4.3.3 Two-stage porosity control results 58 4.4 Discussion 60 4.4.1 Ice crystal growth prediction using ResUnet (6 fold) 64 4.4.2 Dendrite Structure Characteristics 66 4.5 Chapter Summary 69 Chapter 5 Biological Structure Type Control using GAN 70 5.1 Three-dimensional structure generation using generative adversarial networks 70 5.2 Materials and Methods 71 5.2.1 Dataset of biological structures 72 5.2.2 Model architecture of Transformer-Based Conditional GAN 73 5.3 Results 79 5.3.1 The results of the five biological structures 79 5.3.2 Gradient control of biological materials 85 5.4 Discussion 91 5.5 Chapter Summary 93 Chapter 6 Inverse Design of Structural using Deep Learning and Reinforcement Learning 94 6.1 Inverse Design using Structural Properties as Inputs 94 6.2 Materials and Methods 95 6.2.1 Dataset of Triply Periodic Minimal Surfaces 96 6.2.2 Finite element method simulation and automated data generation 97 6.2.3 Structural Generate Models and Generate Framework 99 6.2.4 Inverse Design Framework 102 6.2.5 Conditions Design using Double Deep Q Network (DDQN) 103 6.3 Results 105 6.3.1 Results of 3D structure Generate 105 6.3.2 Results of properties prediction models 108 6.3.3 Results of Generate process 110 6.3.4 Results of Condition Design Using Reinforcement Learning 111 6.4 Discussion 114 6.4.1 Training process of properties prediction model 114 6.4.2 Structure generation for properties outside the dataset 116 6.5 Chapter Summary 119 Chapter 7 Conclusion 120 7.1 Conclusions 120 7.2 Future Work 123 References 124 Appendices 131 Model architecture of ResUnet 131 Gradient control of biological materials 133

    1. Zhang, Q. et al. Bioinspired engineering of honeycomb structure - Using nature to inspire human innovation. Prog. Mater. Sci. 74, 332–400 (2015).
    2. Sun, J., Liu, C., Du, H. &Tong, J. Design of a bionic aviation material based on the microstructure of beetle’s elytra. Int. J. Heat Mass Transf. 114, 62–72 (2017).
    3. Tran, K. T. M. &Nguyen, T. D. Lithography-based methods to manufacture biomaterials at small scales. J. Sci. Adv. Mater. Devices 2, 1–14 (2017).
    4. Zinno, A., Prota, A., DiMaio, E. &Bakis, C. E. Experimental characterization of phenolic-impregnated honeycomb sandwich structures for transportation vehicles. Compos. Struct. 93, 2910–2924 (2011).
    5. Cha, C., Shin, S. R., Annabi, N., Dokmeci, M. R. &Khademhosseini, A. Carbon-based nanomaterials: multifunctional materials for biomedical engineering. ACS Publ. 7, 2891–2897 (2013).
    6. Ragesh, P., Anand Ganesh, V., Nair, S.V. &Nair, A. S. A review on ‘self-cleaning and multifunctional materials’. J. Mater. Chem. A 2, 14773–14797 (2014).
    7. Cobley, C. M., Chen, J., Chul Cho, E., Wang, L.V. &Xia, Y. Gold nanostructures: A class of multifunctional materials for biomedical applications. Chem. Soc. Rev. 40, 44–56 (2011).
    8. Meyers, M. A., Lin, A. Y. M., Chen, P. Y. &Muyco, J. Mechanical strength of abalone nacre: Role of the soft organic layer. J. Mech. Behav. Biomed. Mater. 1, 76–85 (2008).
    9. Ha, N. S., Pham, T. M., Tran, T. T., Hao, H. &Lu, G. Mechanical properties and energy absorption of bio-inspired hierarchical circular honeycomb. Compos. Part B Eng. 236, 109818 (2022).
    10. Li, Y. et al. Simple and efficient volume merging method for triply periodic minimal structures. Comput. Phys. Commun. 264, 107956 (2021).
    11. Tan, T. et al. Mechanical properties of functionally graded hierarchical bamboo structures. Acta Biomater. 7, 3796–3803 (2011).
    12. Michielsen, K. &Stavenga, D. G. Gyroid cuticular structures in butterfly wing scales: biological photonic crystals. J. R. Soc. Interface 5, 85–94 (2007).
    13. Wegst, U. G. K., Bai, H., Saiz, E., Tomsia, A. P. &Ritchie, R. O. Bioinspired structural materials. 14, 23–36 (2015).
    14. Mille, C., Tyrode, E. C. &Corkery, R. W. Inorganic chiral 3-D photonic crystals with bicontinuous gyroid structure replicated from butterfly wing scales. Chem. Commun. 47, 9873–9875 (2011).
    15. Gao, X. et al. Structural and mechanical properties of bamboo fiber bundle and fiber/bundle reinforced composites: a review. J. Mater. Res. Technol. 19, 1162–1190 (2022).
    16. Butler, K. T., Davies, D. W., Cartwright, H., Isayev, O. &Walsh, A. Machine learning for molecular and materials science. Nature vol. 559 547–555 (2018).
    17. Carleo, G. et al. Machine learning and the physical sciences. Rev. Mod. Phys. 91, 045002 (2019).
    18. Yu, C. H., Qin, Z. &Buehler, M. J. Artificial intelligence design algorithm for nanocomposites optimized for shear crack resistance. Nano Futur. 3, (2019).
    19. Yang, Z., Yu, C. H. &Buehler, M. J. Deep learning model to predict complex stress and strain fields in hierarchical composites. Sci. Adv. 7, 1–17 (2021).
    20. Hsu, Y. C., Yu, C. H. &Buehler, M. J. Using Deep Learning to Predict Fracture Patterns in Crystalline Solids. Matter 3, 197–211 (2020).
    21. Lew, A. J., Yu, C. H., Hsu, Y. C. &Buehler, M. J. Deep learning model to predict fracture mechanisms of graphene. npj 2D Mater. Appl. 5, (2021).
    22. Karniadakis, G. E. et al. Physics-informed machine learning. Nat. Rev. Phys. 2021 36 3, 422–440 (2021).
    23. Yu, C. H. et al. Hierarchical Multiresolution Design of Bioinspired Structural Composites Using Progressive Reinforcement Learning. Adv. Theory Simulations 5, 2200459 (2022).
    24. Tseng, B.-Y., Cai, Y.-C., Conan Guo, C.-W., Zhao, E. &Yu, C.-H. Reinforcement learning design framework for nacre-like structures optimized for pre-existing crack resistance. J. Mater. Res. Technol. 24, 3502–3512 (2023).
    25. Tseng, B. Y., Guo, C. W. C., Chien, Y. C., Wang, J. P. &Yu, C. H. Deep Learning Model to Predict Ice Crystal Growth. Adv. Sci. 2207731 (2023).
    26. Hasselt, H.van, Guez, A. &Silver, D. Deep Reinforcement Learning with Double Q-Learning. Proc. AAAI Conf. Artif. Intell. 30, (2016).
    27. Gu, S., Lillicrap, T., Sutskever, I. &Levine, S. Continuous Deep Q-Learning with Model-based Acceleration. in Proceedings of The 33rd International Conference on Machine Learning (eds. Balcan, M. F. &Weinberger, K. Q.) vol. 48 2829–2838 (PMLR, 2016).
    28. He, K., Zhang, X., Ren, S. &Sun, J. Deep residual learning for image recognition. in Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition vols 2016-Decem 770–778 (IEEE Computer Society, 2016).
    29. Ronneberger, O., Fischer, P. &Brox, T. U-net: Convolutional networks for biomedical image segmentation. Lect. Notes Comput. Sci. (including Subser. Lect. Notes Artif. Intell. Lect. Notes Bioinformatics) 9351, 234–241 (2015).
    30. Parmar, N. et al. Image transformer. 35th Int. Conf. Mach. Learn. ICML 2018 9, 6453–6462 (2018).
    31. Vaswani, A. et al. Attention is all you need. Adv. Neural Inf. Process. Syst. 2017 5999–6009 (2017).
    32. Song, Y., Dhariwal, P., Chen, M. &Sutskever, I. Consistency Models. Proc. Mach. Learn. Res. 202, 32211–32252 (2023).
    33. Sui, F., Guo, R., Zhang, Z., Gu, G. X. &Lin, L. Deep Reinforcement Learning for Digital Materials Design. ACS Mater. Lett. 3, 1433–1439 (2021).
    34. Sun, X. et al. Machine learning-enabled forward prediction and inverse design of 4D-printed active plates. Nat. Commun. 15, 5509 (2024).
    35. Deng, B. et al. Inverse Design of Mechanical Metamaterials with Target Nonlinear Response via a Neural Accelerated Evolution Strategy. Adv. Mater. 34, 2206238 (2022).
    36. Zheng, X., Chen, T.Te, Jiang, X., Naito, M. &Watanabe, I. Deep-learning-based inverse design of three-dimensional architected cellular materials with the target porosity and stiffness using voxelized Voronoi lattices. Sci. Technol. Adv. Mater. 24, 2157682 (2023).
    37. Tajmajer, T. Modular multi-objective deep reinforcement learning with decision values. Proc. 2018 Fed. Conf. Comput. Sci. Inf. Syst. FedCSIS 2018 85–93 (2018).
    38. Kobayashi, R. Modeling and numerical simulations of dendritic crystal growth. Phys. D Nonlinear Phenom. 63, 410–423 (1993).
    39. Bahloul, A., Doghri, I. &Adam, L. An enhanced phase field model for the numerical simulation of polymer crystallization. Polym. Cryst. 3, e10144 (2020).
    40. Shibkov, A. A., Golovin, Y. I., Zheltov, M. A., Korolev, A. A. &Leonov, A. A. Morphology diagram of nonequilibrium patterns of ice crystals growing in supercooled water. Phys. A Stat. Mech. its Appl. 319, 65–79 (2003).
    41. Shah, A., Haider, A. &Shah, S. K. Numerical Simulation of Two-Dimensional Dendritic Growth Using Phase-Field Model. World J. Mech. 04, 128–136 (2014).
    42. Lan, C. W., Chang, Y. C. &Shih, C. J. Adaptive phase field simulation of non-isothermal free dendritic growth of a binary alloy. Acta Mater. 51, 1857–1869 (2003).
    43. ZHU, C., LEI, P., XIAO, R. &FENG, L. Phase-field modeling of dendritic growth under forced flow based on adaptive finite element method. Trans. Nonferrous Met. Soc. China 25, 241–248 (2015).
    44. Fabbri, M. &Voller, V. R. The phase-field method in the sharp-interface limit: A comparison between model potentials. J. Comput. Phys. 130, 256–265 (1997).
    45. Libbrecht, K. G. Physical Dynamics of Ice Crystal Growth. Annu. Rev. Mater. Res. 47, 271–295 (2017).
    46. Hu, C., Martin, S. &Dingreville, R. Accelerating phase-field predictions via recurrent neural networks learning the microstructure evolution in latent space. Comput. Methods Appl. Mech. Eng. 397, 115128 (2022).
    47. Yang, K. et al. Self-supervised learning and prediction of microstructure evolution with convolutional recurrent neural networks. Patterns 2, 100243 (2021).
    48. Sharma, A. et al. Optimal Control of Material Micro-Structures. arXiv 2210.06734, (2022).
    49. Peivaste, I. et al. Machine-learning-based surrogate modeling of microstructure evolution using phase-field. Comput. Mater. Sci. 214, 111750 (2022).
    50. Wan, Z., Chang, Z., Xu, Y., Huang, Y. &Šavija, B. Inverse Design of Digital Materials Using Corrected Generative Deep Neural Network and Generative Deep Convolutional Neural Network. Adv. Intell. Syst. 5, 2200333 (2023).
    51. Demange, G., Zapolsky, H., Patte, R. &Brunel, M. A phase field model for snow crystal growth in three dimensions. npj Comput. Mater. 3, 15 (2017).
    52. Demange, G., Zapolsky, H., Patte, R. &Brunel, M. Growth kinetics and morphology of snowflakes in supersaturated atmosphere using a three-dimensional phase-field model. Phys. Rev. E 96, 22803 (2017).
    53. Lee, C. et al. Phase-field computations of anisotropic ice crystal growth on a spherical surface. Comput. Math. with Appl. 125, 25–33 (2022).
    54. Dang, Y., Ai, J., Dai, J., Zhai, C. &Sun, W. Comparative Study on Snowflake Dendrite Solidification Modeling Using a Phase-Field Model and by Cellular Automaton. Processes 10, (2022).
    55. Tan, Q. (谈倩艳), Hosseini, S. A., Seidel-Morgenstern, A., Thévenin, D. &Lorenz, H. Modeling ice crystal growth using the lattice Boltzmann method. Phys. Fluids 34, 13311 (2022).
    56. Pieters, R. &Langer, J. S. Noise-Driven Sidebranching in the Boundary-Layer Model of Dendritic Solidification. Phys. Rev. Lett. 56, 1948–1951 (1986).
    57. Barber, M. N., Barbieri, A. &Langer, J. S. Dynamics of dendritic sidebranching in the two-dimensional symmetric model of solidification. Phys. Rev. A 36, 3340–3349 (1987).
    58. Demange, G., Patte, R. &Zapolsky, H. Induced side-branching in smooth and faceted dendrites: Theory and phase-field simulations. Philos. Trans. R. Soc. A Math. Phys. Eng. Sci. 380, 20200304 (2022).
    59. Karma, A. &Rappel, W. J. Quantitative phase-field modeling of dendritic growth in two and three dimensions. Phys. Rev. E - Stat. Physics, Plasmas, Fluids, Relat. Interdiscip. Top. 57, 4323–4349 (1998).
    60. Jeong, J. H., Goldenfeld, N. &Dantzig, J. A. Phase field model for three-dimensional dendritic growth with fluid flow. Phys. Rev. E - Stat. Physics, Plasmas, Fluids, Relat. Interdiscip. Top. 64, 14 (2001).
    61. Karma, A. &Rappel, W. J. numerical simulation of three-dimensional dendritic growth. Phys. Rev. Lett. 77, 4050–4053 (1996).
    62. Chen, W., Zhao, Y., Yang, S., Zhang, D. &Hou, H. Three-dimensional phase-field simulations of the influence of diffusion interface width on dendritic growth of Fe-0.5 wt.%C alloy. Adv. Compos. Hybrid Mater. 4, 371–378 (2021).
    63. Libbrecht, K. G. An experimental apparatus for observing deterministic structure formation in plate-on-pedestal ice crystal growth. arXiv 1503.01019, (2015).
    64. Chiang, Y.-H. et al. Generating three-dimensional bioinspired microstructures using transformer-based generative adversarial network. J. Mater. Res. Technol. 27, 6117–6134 (2023).

    無法下載圖示 校內：2030-01-14公開
    校外：2030-01-14公開
    電子論文尚未授權公開，紙本請查館藏目錄
    QR CODE