簡易檢索 / 詳目顯示

回結果列表
研究生： 黃柏松
Huang, Po-Sung
論文名稱： 透過緊密連接卷積神經網路開發自動化銑削製程規劃系統
Development of Automatic Milling Process Planning System by Densely Connected Convolutional Neural Networks
指導教授： 鍾俊輝
Chung, Chun-Hui
學位類別： 碩士
Master
系所名稱： 工學院 - 機械工程學系
Department of Mechanical Engineering
論文出版年： 2023
畢業學年度： 111
語文別： 中文
論文頁數： 127
中文關鍵詞： 預訓練模型 、遷移學習 、3D銑削路徑 、刀具路徑參數預測 、CNC加工
外文關鍵詞： Pre-training model, Transfer Learning, Milling path of 3D, Prediction of tool path parameters, CNC
相關次數： 點閱：89  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 致謝 xvi 目錄 xvii 表目錄 xx 圖目錄 xxii 第一章 緒論 1 1.1 研究背景 1 1.2 文獻回顧 2 1.2.1 機器學習輔助工件特徵辨識技術之發展 2 1.2.2 神經網路開發製程刀具路徑規劃之探討 3 1.2.3 人工智慧演進歷程與實務應用驗證概述 6 1.3 研究目的 8 1.4 論文架構 8 第二章 電腦輔助製程規劃 10 2.1 加工工序編排 11 2.2 加工方法選用 13 2.2.1 型腔銑 13 2.2.2 剩餘銑 14 2.2.3 底壁銑 15 2.2.4 深度輪廓銑 15 2.2.5 區域輪廓銑 16 2.2.6 曲面區域輪廓銑 16 2.3 加工策略選用 17 2.3.1 來回往復 17 2.3.2 跟隨周邊 17 2.3.3 跟隨零件 17 2.4 切削刀具種類、尺寸與切削參數之選用 18 2.4.1 端銑刀 19 2.4.2 球銑刀 19 2.4.3 鑽孔刀 20 2.5 零件幾何特徵與尺寸量測 21 2.5.1 加工特徵類別 22 2.5.2 特徵尺寸量測方式 24 2.6 曲面加工方法選用 29 2.7 刀具路徑規劃流程之窮舉方法 39 2.7.1 工序、方法、策略編排規則 40 2.7.2 刀具尺寸選用規則 41 第三章 深層神經網路模型 43 3.1 卷積神經網路模型 43 3.1.1 卷積層與Filter 44 3.1.2 填充方法 45 3.1.3 池化層 46 3.1.4 Dropout 46 3.1.5 激活函數 47 3.1.6 損失函數 49 3.2 遷移學習 50 3.3 預訓練模型 51 3.4 輸入特徵資料蒐集 53 3.4.1 三軸剖面合成視圖 53 3.4.2 Z軸切層視圖 55 3.4.3 特徵尺寸資料 55 3.5 輸出標籤資料蒐集 56 3.6 資料處理 58 3.6.1 資料降維 58 3.6.2 資料增生 59 第四章 銑削製程參數預測模型之建立方法 61 4.1 模型架構 61 4.2 實驗對象為2.5D零件之模型訓練過程 68 4.3 實驗對象為2.5D零件之評估結果與比較 71 4.4 實驗對象為3D零件之模型訓練過程 76 4.4.1 初始完整資料集 78 4.4.2 變更輸入特徵尺寸資料與調整模型結構及迭代次數 81 4.4.3 以DenseNet121最佳架構模型，比對其餘深度學習模型 84 4.5 實驗對象為3D零件之評估結果與比較 92 4.5.1 初始完整資料集 92 4.5.2 變更輸入特徵尺寸資料與調整模型結構及迭代次數 93 4.5.3 以DenseNet121最佳架構模型，比對其餘深度學習模型 98 4.6 實際預測結果與比較 103 第五章 銑削刀具路徑自動化生成系統開發 108 5.1 系統開發架構與環境 108 5.2 系統操作介面與功能 109 第六章 結論與未來展望 111 6.1 結論 111 6.2 未來展望 113 參考文獻 114 附錄 120

    [1] S.Igari, F.Tanaka, M.Onosato, “Computer-aided operation planning for an actual machine tool based on updatable machining database and database-oriented planning algorithm”, Int. J. of Automation Technology, Vol.6, No.6, pp. 717-723, 2012, doi: 10.20965/ijat.2012.p0717.
    [2] E.O.Ezugwu, S.J.Arthur, E. L.Hines, “Tool-wear prediction using artificial neural networks”, J. of Materials Processing Technology, Vol.49, No.3-4, pp.255-265, 1996, doi:10.1016/0924-0136(94)01351-Z.
    [3] G. C. Onwubolu, “Manufacturing features recognition using backpropagation neural networks”, J. of Intelligent Manufacturing, Vol.10, No.3-4, pp.289-299, 1999, doi: 10.1023/a:1008904109029.
    [4] Q. Sun, Z. Yu, Y. Li, S. Yang, J. Xu and H. Yu, “Wear status prediction of micro milling tools by transfer learning and ViT model”, 2021 IEEE International Conference on Manipulation, Manufacturing and Measurement on the Nanoscale (3M-NANO), pp. 183-187, 2021 ,doi: 10.1109/3M-NANO49087.2021.9599807.
    [5] T. Bergs, C. Holst, P. Gupta, and T. Augspurger, “Digital image processing with deep learning for automated cutting tool wear detection,” Procedia Manufacturing, Vol.48, pp. 947-958, 2020, doi: 10.1016/j.promfg.2020.05.134.
    [6] M. M. Isnaini and K. Shirase, “Review of computer-aided process planning systems for machining operation – future development of a computer-aided process planning system”, Int. J. Automation Technology, Vol.5, No.5, pp. 317-332, 2014, doi: 10.20965/ijat.2014.p0317.
    [7] M. M. Isnaini, Y. Shinoki, R. Sato, K. Shirase, “Development of a CAD-CAM interaction system to generate a flexible machining process plan”, Int. J. of Automation Technology, Vol.9, No.2, pp. 104-114, 2015, doi: 10.20965/ijat.2015.p0104.
    [8] E. Morinaga, T. Hara, H. Joko, H. Wakamatsu, E. Arai, “Improvement of computational efficiency in flexible computer-aided process planning”, Int. J. of Automation Technology, Vol.8, No.3, pp. 396-405, 2014, doi: 10.20965/ijat.2014.p0396.
    [9] K. Shirase, K. Nakamoto, “Simulation technologies for the development of an autonomous and intelligent machine tool”, Int. J. of Automation Technology, Vol.7, No.1, pp. 6-15, 2013, doi: 10.20965/ijat.2013.p0006.
    [10] M. Hashimoto, K. Nakamoto, “A Neural Network Based Process Planning System to Infer Tool Path Pattern for Complicated Surface Machining”, Int. J. of Automation Technology, Vol.13, No.1, pp.67-73, 2019, doi: 10.20965/ijat.2019.p0067.
    [11] M. Hashimoto, K. Nakamoto, “Process Planning for Die and Mold Machining Based on Pattern Recognition and Deep Learning”, J. of Advanced Mechanical Design, Systems, and Manufacturing, Vol.15, No.2, JAMDSM0015, 2021, doi: 10.1299/JAMDSM.2021JAMDSM0015.
    [12] O. Cicek, A. Abdulkadir, S. S. Lienkamp, T. Brox, O. Ronneberger, “3d U-Net: Learning dense volumetric segmentation from sparse annotation”, International Conference on Medical Image Computing and Computer-Assisted Intervention – MICCAI 2016, Part II, pp. 424-432, 2016, doi: 10.1007/978-3-319-46723-8_49.
    [13] L. Chen, P. Bender, P. Renton, T. El-Wardany, “Integrated virtual manufacturing systems for process optimization and monitoring”, CIRP Annals- Manufacturing Technology, Vol.51, No.1, pp. 409-412, 2002, doi:10.1016/S0007-8506(07)61548-0.
    [14] Y. Altintas, S. D. Merdol, “Virtual high-performance milling”, CIRP Annals- Manufacturing Technology, Vol.56, No.1, pp.81–84, 2007, doi: 10.1016/j.cirp.2007.05.022.
    [15] T. Deng, Y. Li, X. Liu, P. Wang, K. Lu, “A data-drivenParameter planning method for structural parts NC machining”, Robotics and Computer-Integrated Manufacturing, Vol.68, 2021, Article 102080, doi: 10.1016/j.rcim.2020.102080.
    [16] A. M. Ramos, C. Relvas, J. A. Simoes, “The influence of finishing milling strategies on texture, roughness and dimensional deviations on the machining of complex surfaces”, J. of Materials Processing Technology, Vol.136, No.1-3, pp.209-216, 2003, doi:10.1016/S0924-0136(03)00160-2.
    [17] M. Rybicki, “Problems during milling and roughness registration of free-form surfaces”, J. of Physics: Conference Series, Vol.483, 012007, 2014, doi:10.1088/1742-6596/483/1/012007.
    [18] C. Manav, H. S. Bank, I. Lazoglu, “Intelligent toolpath selection via multi-criteria optimization in complex sculptured surface milling”, J. of Intelligent Manufacturing, Vol.24, No.2, pp. 349-355, 2013, doi: 10.1007/s10845-011-0596-3.
    [19] I. A. Shchurov, L. H. Al-Taie, “Constant scallop-height tool path generation for ball-end mill cutters and three-axis CNC milling machines”. Procedia Engineering, Vol.206, pp.1137-1141, 2017, doi: 10.1016/j.proeng.2017.10.607.
    [20] R. Q. Sardinas, P. Reis, J. P. Davim, “Multi-objective optimization of cutting parameters for drilling laminate composite materials by using genetic algorithms”, Composites Science and Technology, Vol.66, No.15, pp.3083–3088, 2006, doi: 10.1016/j.compscitech.2006.05.003.
    [21] 李佳政, “以NX開發被夾持零件之自動刀具路徑規劃系統”, 國立台灣科技大學機械工程系碩士學位論文, 2019.
    [22] J. Xu, Z. Geng, Y. Sun, L. Li, “Generating gouge-free tool paths for ball-end cutter CNC milling of cloud of point by projecting guide curves”, Int. J. of Advanced Manufacturing Technology, Vol.102, No.5-8, pp. 1193–1204 , 2019, doi:10.1007/s00170-018-2973-y.
    [23] 張宏銘, “使用卷積神經網路開發2.5D零件自動化銑削製程規劃系統”, 國立成功大學機械工程學系碩士論文, 2022.
    [24] L.T. Tunc, “Smart tool path generation for 5-axis ball-end milling of sculptured surfaces using process models”, Robotics and Computer-Integrated Manufacturing, Vol.56, pp. 212-221, 2019, doi:10.1016/j.rcim.2018.10.002.
    [25] M. A. Dittrich, F. Uhlich, B. Denkena, “Self-optimizing tool path generation for 5-axis machining processes”, CIRP J. of Manufacturing Science and Technology, Vol.24, pp. 49-54, 2019, doi:10.1016/j.cirpj.2018.11.005.
    [26] M. Iman, J. A. Miller, K. Rasheed, R. M. Branch, H. R. Arabnia, “EXPANSE: A Deep Continual / Progressive Learning System for Deep Transfer Learning”, 2022 International Conference on Computational Science and Computational Intelligence (CSCI 2022), 2022, doi:10.48550/arXiv.2205.10356.
    [27] B. Denkena, M. -A. Dittrich, F. Uhlich, “Augmenting Milling Process Data for Shape Error Prediction”. Procedia CIRP, Vol.57, pp.487-491, 2016, doi:10.1016/j.procir.2016.11.084.
    [28] C. Finn, P. Abbeel, S. Levine, “Model-agnostic meta learning for fast adaptation of deep networks”, 34th International Conference on Machine Learning (ICML), PMLR, Vol.70, pp.1126-1135, 2017.
    [29] T. Munkhdalai, H. Yu, “Meta networks”, 34th International Conference on Machine Learning (ICML), PMLR, Vol.70, pp. 2554-2563, 2017.
    [30] S. N. Venkatesh, P. A. Balaji, M. Elangovan, K. Annamalai, V.Indira, V.Sugumaran, V.S. Mahamuni, “Transfer Learning-Based Condition Monitoring of Single Point Cutting Tool”, Computational Intelligence and Neuroscience, Vol.2022, 2022, doi: 10.1155/2022/3205960.
    [31] J. Xie, P. Hu, S. Gao, J. Chen, H. Zhou, J. Yang, “Efficient Cutting Power Modeling of Three-Axis Milling Based on Transfer Learning and Neural Network”, IEEE Access, vol. 10, pp. 64451-64464, 2022, doi: 10.1109/ACCESS.2022.3184023.
    [32] A. A. Rusu, N. C. Rabinowitz, G. Desjardins, H. Soyer, J. Kirkpatrick, K. Kavukcuoglu, R. Pascanu, R. Hadsell, “Progressive neural networks”, preprint arXiv:1606.04671, 2016.
    [33] L. T. Tunç, O. M. Ozkirimli, E. Budak, “Machining strategy development and parameter selection in 5-axis milling based on process simulations”, Int. J. of Advanced Manufacturing Technology, Vol.85, No.5-8, pp.1483-1500, 2016, doi: 10.1007/s00170-015-8001-6.
    [34] 震虎精密科技股份有限公司. “震虎綜合型錄General Catalogue.” https://www.speedtiger.com.tw/document/
    [35] 震虎精密科技股份有限公司. “鎢鋼鑽頭型錄Carbide Drills Catalogue.” https://www.speedtiger.com.tw/document/
    [36] 陳家銘, “使用動態規劃法進行2.5D銑削刀具路徑及餘量設定之最佳化”, 國立成功大學機械工程學系碩士論文, 2020.
    [37] 陳亮頤, “以基因演算法進行2.5D工件銑削刀具路徑規劃及餘量設定之最佳化”, 國立成功大學機械工程學系碩士論文, 2022.
    [38] D. M. D’addona, R. Teti, “Genetic algorithm-based optimization of cutting parameters in turning processes”, Procedia CIRP, Vol.7, pp.323-328, 2013, doi: 10.1016/j.procir.2013.05.055.
    [39] Y. Lecun, L. Bottou, Y. Bengio, P. Haffner, “Gradient-based learning applied to document recognition”, Proceedings of the IEEE, vol.86, no.11, pp.2278-2324, 1998, doi: 10.1109/5.726791.
    [40] A. Krizhevsky, I. Sutskever, G. E. Hinton, “Imagenet classification with deep convolutional neural networks”, 25th International Conference on Neural Information Processing Systems (NIPS), pp.1097-1105, 2012.
    [41] K. Simonyan, A. Zisserman, “Very Deep Convolutional Networks for Large-Scale Image Recognition”, 3rd International Conference on Learning Representations (ICLR), pp.1-14, 2015.
    [42] C. Szegedy et al., “Going deeper with convolutions”, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 1-9, 2015, doi: 10.1109/CVPR.2015.7298594.
    [43] K. He, X. Zhang, S. Ren, J. Sun, “Deep residual learning for image recognition”, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 770-778, 2016, doi: 10.1109/CVPR.2016.90.
    [44] N. Ozturk, F. Ozturk, “Neural network based non-standard feature recognition to integrate CAD and CAM”, Computers in Industry, Vol.45, No.2, pp.123-135, 2001, doi: 10.1016/S0166-3615(01)00090-2.
    [45] Keras, (n.d.), Available models, Keras Applications, https://keras.io/api/applications/.
    [46] G. Huang, Z. Liu, L. Van Der Maaten, K. Q. Weinberger, “Densely Connected Convolutional Networks”, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 4700-4708, 2017.

    無法下載圖示 校內：2028-08-11公開
    校外：2028-08-11公開
    電子論文尚未授權公開，紙本請查館藏目錄
    QR CODE