Skip navigation

DSpace

機構典藏 DSpace 系統致力於保存各式數位資料(如:文字、圖片、PDF)並使其易於取用。

點此認識 DSpace
DSpace logo
English
中文
  • 瀏覽論文
    • 校院系所
    • 出版年
    • 作者
    • 標題
    • 關鍵字
  • 搜尋 TDR
  • 授權 Q&A
    • 我的頁面
    • 接受 E-mail 通知
    • 編輯個人資料
  1. NTU Theses and Dissertations Repository
  2. 工學院
  3. 化學工程學系
請用此 Handle URI 來引用此文件: http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/86323
完整後設資料紀錄
DC 欄位值語言
dc.contributor.advisor郭修伯(Hsiu-Po Kuo)
dc.contributor.authorChia-You Liuen
dc.contributor.author劉家佑zh_TW
dc.date.accessioned2023-03-19T23:49:04Z-
dc.date.copyright2022-09-02
dc.date.issued2022
dc.date.submitted2022-08-26
dc.identifier.citation1. Raissi, M., P. Perdikaris, and G.E. Karniadakis, Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. Journal of Computational Physics, 2019. 378: p. 686-707. 2. Williams, J.C., The mixing of dry powders. Powder Technology, 1968. 2(1): p. 13-20. 3. Bridgwater, J., Fundamental powder mixing mechanisms. Powder Technology, 1976. 15(2): p. 215-236. 4. Kuo, H., R. Hsu, and Y. Hsiao, Investigation of axial segregation in a rotating drum. Powder technology, 2005. 153(3): p. 196-203. 5. Cundall, P.A. and O.D. Strack, A discrete numerical model for granular assemblies. geotechnique, 1979. 29(1): p. 47-65. 6. Park, J. and N. Kang, Applications of fiber models based on discrete element method to string vibration. Journal of Mechanical Science and Technology, 2009. 23(2): p. 372-380. 7. Dury, C.M. and G.H. Ristow, Competition of mixing and segregation in rotating cylinders. Physics of fluids, 1999. 11(6): p. 1387-1394. 8. Hlungwani, O., et al., Further validation of DEM modeling of milling: effects of liner profile and mill speed. Minerals Engineering, 2003. 16(10): p. 993-998. 9. Xu, Y., et al., 2D DEM simulation of particle mixing in rotating drum: A parametric study. Particuology, 2010. 8(2): p. 141-149. 10. Yamamoto, M., S. Ishihara, and J. Kano, Evaluation of particle density effect for mixing behavior in a rotating drum mixer by DEM simulation. Advanced Powder Technology, 2016. 27(3): p. 864-870. 11. Nishiura, D. and H. Sakaguchi, Parallel-vector algorithms for particle simulations on shared-memory multiprocessors. Journal of Computational Physics, 2011. 230(5): p. 1923-1938. 12. Xu, J., et al., Quasi-real-time simulation of rotating drum using discrete element method with parallel GPU computing. Particuology, 2011. 9(4): p. 446-450. 13. Gan, J., Z. Zhou, and A. Yu, A GPU-based DEM approach for modelling of particulate systems. Powder Technology, 2016. 301: p. 1172-1182. 14. Lu, L., et al., Computer virtual experiment on fluidized beds using a coarse-grained discrete particle method—EMMS-DPM. Chemical Engineering Science, 2016. 155: p. 314-337. 15. Sakai, M., et al., Verification and validation of a coarse grain model of the DEM in a bubbling fluidized bed. Chemical Engineering Journal, 2014. 244: p. 33-43. 16. Di Renzo, A., E.S. Napolitano, and F.P. Di Maio, Coarse-grain dem modelling in fluidized bed simulation: A review. Processes, 2021. 9(2): p. 279. 17. Andrews, M.J. and P.J. O'Rourke, The multiphase particle-in-cell (MP-PIC) method for dense particulate flows. International Journal of Multiphase Flow, 1996. 22(2): p. 379-402. 18. Rosenblatt, F., The perceptron: a probabilistic model for information storage and organization in the brain. Psychological review, 1958. 65(6): p. 386. 19. Minsky, M. and S. Papert, An introduction to computational geometry. Cambridge tiass., HIT, 1969. 479: p. 480. 20. Rumelhart, D.E., G.E. Hinton, and R.J. Williams, Learning representations by back-propagating errors. nature, 1986. 323(6088): p. 533-536. 21. Hornik, K., M. Stinchcombe, and H. White, Multilayer feedforward networks are universal approximators. Neural networks, 1989. 2(5): p. 359-366. 22. Krizhevsky, A., I. Sutskever, and G.E. Hinton, Imagenet classification with deep convolutional neural networks. Advances in neural information processing systems, 2012. 25. 23. Hochreiter, S. and J. Schmidhuber, Long short-term memory. Neural computation, 1997. 9(8): p. 1735-1780. 24. Kolmogorov, A.N., The local structure of turbulence in incompressible viscous fluid for very large Reynolds numbers. Cr Acad. Sci. URSS, 1941. 30: p. 301-305. 25. Kutler, P. and U. Mehta. Computational aerodynamics and artificial intelligence. in 17th Fluid Dynamics, Plasma Dynamics, and Lasers Conference. 1984. 26. Teo, C., et al. A neural net approach in analyzing photograph in PIV. in Conference Proceedings 1991 IEEE International Conference on Systems, Man, and Cybernetics. 1991. IEEE. 27. Grant, I. and X. Pan, An investigation of the performance of multi layer, neural networks applied to the analysis of PIV images. Experiments in Fluids, 1995. 19(3): p. 159-166. 28. Bishop, C.M. and G.D. James, Analysis of multiphase flows using dual-energy gamma densitometry and neural networks. Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, 1993. 327(2-3): p. 580-593. 29. Milano, M. and P. Koumoutsakos, Neural network modeling for near wall turbulent flow. Journal of Computational Physics, 2002. 182(1): p. 1-26. 30. Benvenuti, L., C. Kloss, and S. Pirker, Identification of DEM simulation parameters by Artificial Neural Networks and bulk experiments. Powder technology, 2016. 291: p. 456-465. 31. Ye, F., et al., Calibration and verification of DEM parameters for dynamic particle flow conditions using a backpropagation neural network. Advanced Powder Technology, 2019. 30(2): p. 292-301. 32. Ma, C., et al., Calibration of the microparameters of the discrete element method using a relevance vector machine and its application to rockfill materials. Advances in Engineering Software, 2020. 147: p. 102833. 33. Ling, J., A. Kurzawski, and J. Templeton, Reynolds averaged turbulence modelling using deep neural networks with embedded invariance. Journal of Fluid Mechanics, 2016. 807: p. 155-166. 34. He, L. and D.K. Tafti, A supervised machine learning approach for predicting variable drag forces on spherical particles in suspension. Powder technology, 2019. 345: p. 379-389. 35. Jiang, Y., et al., Neural-network-based filtered drag model for gas-particle flows. Powder Technology, 2019. 346: p. 403-413. 36. Ladický, L.u., et al., Data-driven fluid simulations using regression forests. ACM Transactions on Graphics (TOG), 2015. 34(6): p. 1-9. 37. Ummenhofer, B., et al. Lagrangian fluid simulation with continuous convolutions. in International Conference on Learning Representations. 2019. 38. Kochkov, D., et al., Machine learning–accelerated computational fluid dynamics. Proceedings of the National Academy of Sciences, 2021. 118(21). 39. Wiewel, S., M. Becher, and N. Thuerey. Latent space physics: Towards learning the temporal evolution of fluid flow. in Computer graphics forum. 2019. Wiley Online Library. 40. Xie, Y., et al., tempogan: A temporally coherent, volumetric gan for super-resolution fluid flow. ACM Transactions on Graphics (TOG), 2018. 37(4): p. 1-15. 41. Liao, Z., et al., Image-based prediction of granular flow behaviors in a wedge-shaped hopper by combing DEM and deep learning methods. Powder Technology, 2021. 383: p. 159-166. 42. 王萱, 以實驗及模擬探討具有粒子軸向運動擋板之滾動鼓的粒子混合行為. 2020, 長庚大學. 43. Hertz, H., On the contact of rigid elastic solids and on hardness. Hertz, H.; Schott, JA., editors. 1882, London: MacMillan. 44. Mindlin, R.D. and H. Deresiewicz, Elastic spheres in contact under varying oblique forces. 1953. 45. Tsuji, Y., T. Tanaka, and T. Ishida, Lagrangian numerical simulation of plug flow of cohesionless particles in a horizontal pipe. Powder technology, 1992. 71(3): p. 239-250. 46. Sakaguchi, H., E. Ozaki, and T. Igarashi, Plugging of the flow of granular materials during the discharge from a silo. International Journal of Modern Physics B, 1993. 7(09n10): p. 1949-1963. 47. McCulloch, W.S. and W. Pitts, A logical calculus of the ideas immanent in nervous activity. The bulletin of mathematical biophysics, 1943. 5(4): p. 115-133. 48. Han, J. and C. Moraga. The influence of the sigmoid function parameters on the speed of backpropagation learning. in International workshop on artificial neural networks. 1995. Springer. 49. Nair, V. and G.E. Hinton. Rectified linear units improve restricted boltzmann machines. in Icml. 2010. 50. Karpatne, A., et al., Physics-guided neural networks (pgnn): An application in lake temperature modeling. arXiv preprint arXiv:1710.11431, 2017. 51. Sutskever, I., et al. On the importance of initialization and momentum in deep learning. in International conference on machine learning. 2013. PMLR. 52. Duchi, J., E. Hazan, and Y. Singer, Adaptive subgradient methods for online learning and stochastic optimization. Journal of machine learning research, 2011. 12(7). 53. Hinton, G., N. Srivastava, and K. Swersky, Neural networks for machine learning lecture 6a overview of mini-batch gradient descent. Cited on, 2012. 14(8): p. 2. 54. Kingma, D.P. and J. Ba, Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980, 2014. 55. Hochreiter, S., et al., Gradient flow in recurrent nets: the difficulty of learning long-term dependencies. 2001, A field guide to dynamical recurrent neural networks. IEEE Press. 56. Greff, K., et al., LSTM: A search space odyssey. IEEE transactions on neural networks and learning systems, 2016. 28(10): p. 2222-2232. 57. Xu, D. and Y. Shen, An improved machine learning approach for predicting granular flows. Chemical Engineering Journal, 2022: p. 138036. 58. Abadi, M., et al. Tensorflow: A system for large-scale machine learning. in 12th {USENIX} symposium on operating systems design and implementation ({OSDI} 16). 2016. 59. Raissi, M., P. Perdikaris, and G.E. Karniadakis, Multistep neural networks for data-driven discovery of nonlinear dynamical systems. arXiv preprint arXiv:1801.01236, 2018. 60. Lu, L., et al., Machine Learning Accelerated Discrete Element Modeling of Granular Flows. Chemical Engineering Science, 2021: p. 116832.
dc.identifier.urihttp://tdr.lib.ntu.edu.tw/jspui/handle/123456789/86323-
dc.description.abstract離散元素法基於計算粒子之間的交互作用力,進而能準確地模擬出滾動鼓中粒子之大小顆粒粒子運動的偏析行為,但卻也需要龐大的計算成本。在本研究中,我們以三維(3D)與二維(2D)滾動鼓中,以離散元素法所得出的顆粒運動模擬數據作為機器學習的資料庫,以單顆粒子輸入與多顆粒子輸入至ANN、RNN、CNN三種不同的機器學習模型進行訓練。 在單顆顆粒運動訓練上,本研究以3D滾動鼓中的顆粒軌跡進行ANN模型訓練。在粒子徑向移動軌跡的時間外插預測上,得到決定係數0.93的運動預測;於軸向方向大小粒子軌跡預測上,得到決定係數0.65的運動預測。以2D滾動鼓中的顆粒軌跡進行RNN模型訓練,在移動軌跡的時間外插預測使用多對多模型,當參考點達30步時間步時,模型的外插預測能力會開始下降;而在移動軌跡的時間外插預測使用多對一模型時,當參考點增加時,模型的外插能力也跟著增加。 在多顆顆粒運動訓練上,本研究以多顆顆粒時間維度與空間維度上進行了ANN、RNN、CNN等模型之訓練,並以粒子重疊次數作為物理不一致性的比較標準,比較各模型於短時間與長時間的外插預測比較。在外插時間5秒後,ANN模型於物理不一致性上高出RNN與CNN模型分別有8倍與10倍大的誤差量,而RNN模型在30秒外插時間後誤差量也超出CNN模型2倍之多。若要在物理一致性上有更好的表現,需要多加考慮到模型學習空間維度上資訊能力,才可以有較好的預測表現。zh_TW
dc.description.abstractBinary particle interaction based Discrete Element Method (DEM) simulations can accurately predict particle size segregation in a rotating drum. Although DEM simulations are accurate and reliable, they are computational expensive. In this study, we use DEM simulation predicted particle trajectories in a rotating drum as the database for Machine Learning (ML) particle trajectory forecasting training and validation. With single particle information as the input data, we used ANN model and RNN model to learn and predict particle trajectories in a 3D rotating drum and in a 2D rotating drum, respectively. We also used ANN, RNN and CNN machine learning models to learn and predict particle segregation trajectories in a 2D rotating drum by using whole particle information in the drum a the input data. Using the single particle training, the ANN model predicts the trajectories of 5 representative particles with the coefficient of determination R2 of 0.93 in the radial direction and 0.65 in the axial direction in the 5 sec extrapolation time. The RNN model shows that the accuracy of the MtM model design prediction is decreasing when the number of time steps is over 30; for MtO model design, the more number of time steps, the more accrual the model is. Using the whole particles training, because the ANN model doesn’t consider the temporal and spatial information, ANN has more than 8 times over RNN’s number of overlapping, and 10 times of CNN’s at 5 sec extrapolation time. As the extrapolation time increasing, number of particle overlapping increasing. For 30 sec extrapolation time, RNN has more than twice over CNN’s number of overlapping. To improve the model performances, the models should have the ability in learning spatial information when predicting particle size segregation trajectories in 2D rotating drum.en
dc.description.provenanceMade available in DSpace on 2023-03-19T23:49:04Z (GMT). No. of bitstreams: 1
U0001-2508202213070700.pdf: 11881049 bytes, checksum: d90a65670c1435f5d1020031b578e3bb (MD5)
Previous issue date: 2022
en
dc.description.tableofcontents目錄 i 圖表目錄 iii 第一章 緒論 1 第二章 文獻回顧 2 2.1 粒子偏析(PARTICLE SEGREGATION) 2 2.2 離散元素法(DISCRETE ELEMENT METHOD) 3 2.3 機器學習(MACHINE LEARNING) 8 第三章 實驗方法 17 3.1 EDEM軟體粒子運動之參數設定 17 3.1.1 粒子材料與裝置材料性質設定 19 3.1.2 粒子材料與裝置材料尺寸設定 19 3.1.3 材料間作用力設定 20 3.1.4 設定粒子添加與裝置轉速情形 20 3.1.5 設定材料間的物理交互作用 21 3.1.6 設定時間步 23 3.1.7 設定網格大小 23 3.1.8 模擬結果輸出 24 3.2 人工神經網路(ARTIFICIAL NEURAL NETWORKS) 25 3.2.1 正向傳播(Forward Propagation) 25 3.2.2 激活函數(Activation function) 26 3.2.3 損失函數(Loss function) 26 3.2.4 優化器(Optimizer) 28 3.2.5 遞迴神經網路(Recurrent Neural Networks, RNNs) 28 3.2.6 卷積神經網路(Convolutional Neural Networks, CNNs) 30 3.2.7 運算環境 31 3.2.8 資料預處理 31 3.2.9 訓練設定 32 第四章 結果與討論 34 4.1 使用單顆粒子訓練之神經網路結果 34 4.1.1 人工神經網路預測結果 34 4.1.2 遞迴神經網路預測結果 46 4.2 使用多顆粒子訓練之神經網路結果 56 4.2.1 人工神經網路預測結果 56 4.2.2 遞迴神經網路預測結果 59 4.2.3 卷積神經網路預測結果 63 4.2.3.1 超參數調整 65 4.2.3.2 外插預測與模擬比較 74 4.2.3 模型預測能力與物理不一致性之比較 76 4.2.4 模型預測能力驗證 79 第五章 結論 83 參考文獻 85
dc.language.isozh-TW
dc.title以機器學習方式預測粒子於滾動鼓中偏析軌跡zh_TW
dc.titlePrediction of Particle Segregation Trajectories in Rotating Drum by Using Machine Learning Method.en
dc.typeThesis
dc.date.schoolyear110-2
dc.description.degree碩士
dc.contributor.oralexamcommittee徐振哲(Cheng-Che Hsu),楊延齡(Yan-Ling Yang)
dc.subject.keyword離散元素法,機器學習,滾動鼓,粒子偏析,zh_TW
dc.subject.keywordDiscrete Element Method,Machine Learning,Rotating Drum,Particle Segregation,en
dc.relation.page88
dc.identifier.doi10.6342/NTU202202801
dc.rights.note同意授權(全球公開)
dc.date.accepted2022-08-26
dc.contributor.author-college工學院zh_TW
dc.contributor.author-dept化學工程學研究所zh_TW
dc.date.embargo-lift2025-07-31-
顯示於系所單位:化學工程學系

文件中的檔案:
檔案 大小格式 
U0001-2508202213070700.pdf
  此日期後於網路公開 2025-07-31
11.6 MBAdobe PDF
顯示文件簡單紀錄


系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。

社群連結
聯絡資訊
10617臺北市大安區羅斯福路四段1號
No.1 Sec.4, Roosevelt Rd., Taipei, Taiwan, R.O.C. 106
Tel: (02)33662353
Email: ntuetds@ntu.edu.tw
意見箱
相關連結
館藏目錄
國內圖書館整合查詢 MetaCat
臺大學術典藏 NTU Scholars
臺大圖書館數位典藏館
本站聲明
© NTU Library All Rights Reserved