研究生: |
李昭彣 Lee, Chao-Wen |
---|---|
論文名稱: |
以徑向基底函數類神經網路於開放式虛擬工具機系統進行切削力預測之研究 Studies on Cutting Force Prediction with Radial Basis Function Neural Networks in Open-Architecture Virtual Machine Tool System |
指導教授: |
李榮顯
Lee, Rong-Shean |
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: | 2018 |
畢業學年度: | 106 |
語文別: | 中文 |
論文頁數: | 104 |
中文關鍵詞: | 徑向基函數 、類神經網路 、切削力預測 、虛擬工具機 、通訊連線 |
外文關鍵詞: | RBFNN, cutting force prediction, virtual machine tool, artificial neural networks |
相關次數: | 點閱:118 下載:2 |
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 |
五軸工具機為高精度加工帶來了便利,然而隨著加工外型日趨複雜,加工過程中所產生的誤差和危險,更加不可預期。為確保加工過程順利,往往必須反覆測試程式並採取相對保守的加工條件,在時間成本上花費甚鉅。市面上常見的商用加工驗證軟體如NX及Vericut,需耗費額外成本製作不同控制器廠牌或型號的CNC控制器模擬器來進行驗證。
本文發展出將類神經網路結合開放式虛擬工具機的系統,此系統具有資料分析的能力。透過網路通訊建立商用控制器與虛擬工具機間的通訊協定,藉由路徑插補解譯模擬真實加工情況,包含運動模擬、切削模擬、碰撞偵測等。模擬所獲得的切削幾何資訊與從控制器取得的加工資訊匯入類神經網路所建立的模型得以預測結果,此預測結果能提供使用者進行分析。本文同時設計友善之人機介面,方便操作者建立預測模型,調整模型參數與前處理資料,並以圖表化方式呈現結果。虛擬加工環境則以MVC(Model-View-Controller)架構設計,建立仿照真實加工環境之系統。
最後本文透過與不同商用控制器通訊,以鞋模三軸加工案例與葉輪五軸加工案例呈現模擬結果,並比較類神經網路預測模型所得預測切削力與物理力學模型計算切削力的結果與趨勢。
The five-axis machine tool brings performance to high-precision manufacturing. However, with complex workpieces unpredictable errors and dangers during the manufacturing process increase. In order to ensure the success of the manufacturing process, it’s necessary to repeat tests and adopt more conservative processing strategies, which leads to a great expense on machining time.
This research develops a system that combines artificial neural networks and an open virtual machine tool, which could analyze process data. By establishing the communication protocol between the commercial controller and the virtual machine tool via network communication, which can receive path interpolation from controller and simulate machining process, including motion simulation of the machine tool, cutting simulation, and collision detection. The geometry data by simulating and manufacturing process information from controller are taken as the inputs import into the neural network. It can construct the predictive model. Besides, human-machine interface has been designed in this thesis, which is convenient for the end-user to set up models, adjust parameters, pre-process data, and presents the results. The virtual machining environment is designed base on MVC architecture to create realistic circumstances.
Finally, this research creates the communication with different commercial controllers, the simulation results are presented in the case of shoe mold in 3-axis machining tool and the case of the wheel rim in 5-axis machining. The research also shows graphic results which the cutting force between the neural network prediction model and calculated physical model can be compared.
Altintas, Y., Brecher, C., Weck, M., and Witt, S.: Virtual machine tool, CIRP Annals-manufacturing technology, 54, 115-138, 2005.
Altintas, Y. and Engin, S.: Generalized modeling of mechanics and dynamics of milling cutters, CIRP Annals-Manufacturing Technology, 50, 25-30, 2001.
Altintaş, Y. and Lee, P.: A general mechanics and dynamics model for helical end mills, CIRP annals, 45, 59-64, 1996.
Atluru, S. and Deshpande, A.: Data to information: can MTConnect deliver the promise, Transactions of NAMRI/SME, 37, 197-204, 2009.
Briceno, J. F., El-Mounayri, H., and Mukhopadhyay, S.: Selecting an artificial neural network for efficient modeling and accurate simulation of the milling process, International Journal of Machine Tools and Manufacture, 42, 663-674, 2002.
Budak, E., Altintas, Y., and Armarego, E.: Prediction of milling force coefficients from orthogonal cutting data, Journal of Manufacturing Science and Engineering, 118, 216-224, 1996.
Ding, S., Mannan, M., and Poo, A. N.: Oriented bounding box and octree based global interference detection in 5-axis machining of free-form surfaces, Computer-Aided Design, 36, 1281-1294, 2004.
Ferry, W. and Altintas, Y.: Virtual five-axis flank milling of jet engine impellers—Part I: mechanics of five-axis flank milling, Journal of Manufacturing Science and Engineering, 130, 011005, 2008.
Heaton, J.: Encog: Library of Interchangeable Machine Learning Models for Java and C#, Journal of Machine Learning Research, 16, 1243-1247, 2015.
Karunakaran, K., Shringi, R., Ramamurthi, D., and Hariharan, C.: Octree-based NC simulation system for optimization of feed rate in milling using instantaneous force model, The International Journal of Advanced Manufacturing Technology, 46, 465-490, 2010.
Koren, Y., Hu, S., Gu, P., and Shpitalni, M.: Open-architecture products, CIRP Annals, 62, 719-729, 2013.
Lee, R. S. and Mei, K. J.: Development of Parallel Cutting Simulation with Adaptive Octree Model in Virtual Machine Tool, The 2012 IEEE/ASME International Conference on Advanced Intelligent Mechatronics, 2012. PP. 654-658, 2012.
Mei, K.-J. and Lee, R.-S.: Collision detection for virtual machine tools and virtual robot arms using the Shared Triangles Extended Octrees method, International Journal of Computer Integrated Manufacturing, 29, 355-373, 2016.
Reenskaug, T.: Models-views-controllers, Technical note, Xerox PARC, 32, 6.2, 1979.
Tandon, V. and El-Mounayri, H.: A novel artificial neural networks force model for end milling, The International Journal of Advanced Manufacturing Technology, 18, 693-700, 2001.
Torrisi, N. M. and de Oliveira, J. F. G.: Remote monitoring for high-speed CNC processes over public IP networks using CyberOPC, The International Journal of Advanced Manufacturing Technology, 60, 191-200, 2012.
Zhang, Y., Xu, X., and Liu, Y.: Numerical control machining simulation: a comprehensive survey, International Journal of Computer Integrated Manufacturing, 24, 593-609, 2011.
Zuperl, U. and Cus, F.: Tool cutting force modeling in ball-end milling using multilevel perceptron, Journal of materials processing technology, 153, 268-275, 2004.
Zuperl, U., Cus, F., Mursec, B., and Ploj, T.: A generalized neural network model of ball-end milling force system, Journal of materials processing technology, 175, 98-108, 2006.
朱育柱: 開放式架構五軸虛擬工具機之整合人機介面開發, 國立成功大學機械所碩士論文, 2016.
吳長銘: 虛擬工具機之開放式建構及在圓弧與轉角誤差改善之應用,, 國立成功大學機械所碩士論文, 2015.
徐鵬: 基於類神經網路切削力預測模型的五軸铣削製程優化之研究, 國立成功大學機械所碩士論文, 2015.
梅可人: 以創新延伸式八元樹法應用於數控加工模擬系統之碰撞檢測與幾何移除, 國立成功大學博士論文, 2016.
黃玲真: 薄工件五軸虛擬加工其彈性變形之體積誤差分析與補償, 國立成功大學機械所碩士論文, 2013.
新代: Syntec Remote API 2015.
楊承融: 應用遺傳演算法於五軸虛擬工具機體積誤差分配, 國立成功大學機械所碩士論文, 2011.
謝恩平: 應用八元樹法於銑削力預估之五軸虛擬工具機系統, 國立成功大學機械所碩士論文, 2013.