臺灣博碩士論文加值系統

刀具於切削時常伴隨著積屑或磨耗等情形發生，其刀具的狀態將會影響其工件的精度以及品質，若是於切削過程中發生撞機或過切等情形，除了刀具可能發生破損外，機台也可能因為撞機而造成損害，進而增加維修的成本，故一個即時性的刀具監測系統開發是必要且重要的。
本研究之刀具監測系統主要以類神經網路為主要演算法，運用本實驗室自行開發之Smart Machine Engine智能製造系統搭配C#進行系統開發，利用OPC UA通訊傳輸，將本研究之系統與CNC工具機進行連結，目的為即時將當前刀具狀態或機台狀態呈現給使用者，降低因狀態異常對機台的損害以及方便後續即時刀具補正系統的開發。本系統之開發主要分為三個部分，依序為「切削監測系統」、「防碰撞系統」以及「刀具磨耗預測系統」。本研究主要利用切削時的主軸負載進行訊號分析，透過一完整工序的加工，監測其切削時相關物理量的變化，進而從中分析出切削時的主軸負載，建置出相應的切削防碰撞模型，後續並運用其負載相關特徵值進行未知磨耗量的預測，以利後續磨耗狀態的掌握。
經由本系統防碰撞模型的建立，除了能即時監測刀具或機台的狀態是否發生異常外，也能降低因異常狀態所帶來之損傷。刀具磨耗預測方面，經由本研究之驗證，其系統磨耗預測值與實際值相差約4 μm，但以其刀長設定器之重複精度誤差6 μm而言，其預測值仍在磨耗之預期精度內，故此預測系統將有利於後續刀具即時補證之開發。

In the cutting process, the tool status will influence the accuracy and quality of workpiece. If the collision or overcut happened, it will damage the tool and the machine with increasing the cost of maintenance. Therefore, developing the tool monitoring system is necessary and essential.
By using the concept of the mechanics, this paper analyzed the signal from the spindle load to create the collision avoidance model and explored its feature to predict the tool wear so that it can handle all of the situation. The purpose of the study was to develop the tool monitoring system for making sure the status of the machine and its tool. For avoiding the abnormal status of the machine and estimating the tool wear, users could create the mathematical module in real time by the cutting load in order to predict the wear. Neural Network and Smart Machine Engine (SME) are the main application of mathematical method and system in this paper. Smart Machine Engine developed by National Formosa University is an intelligent manufacturing management system. This study used the SME system and Visual Studio C# to establish the tool monitoring system. Using the OPC UA communication protocol to connect the tool monitoring system with the machine tool in order to display tool status for users, which can decrease the damage of the abnormal status.

摘要...i
Abstract...iii
誌謝...v
目錄...vi
表目錄...ix
圖目錄...xii
符號說明...xix
第一章 緒論...1
1.1 研究背景與動機...1
1.2 研究目的 ...3
1.3 研究流程 ...4
1.4 國內外相關文獻...6
1.4.1 感測器訊號分析相關文獻...6
1.4.2 切削參數及切削條件分析相關文獻...20
1.4.3 文獻回顧總結...23
1.5 論文架構...23
第二章 系統架構與說明...25
2.1 系統架構 ...25
2.2 Smart Machine Engine系統架構...27
2.3 硬體架構 ...32
2.3.1 臥式車削加工中心...32
2.3.2 刀長設定器...35
2.3.3 電子元件 ...39
2.3.4 雙束型聚焦離子束顯微系統...43
第三章 刀具監測系統建置與原理... 44
3.1 加工工序建立...44
3.1.1 防碰撞系統之加工工序規劃...45
3.1.2 刀具磨耗預測系統之加工工序規劃...47
3.2 Smart Machine Engine連線設定...49
3.3 機器學習(Machine Learning) [49]...50
3.3.1 監督式學習 [52]...52
3.3.2 非監督式學習 [52]...53
3.3.3 增強學習 [52]...53
3.4 刀具狀態監測模型建置...54
3.4.1 類神經網路 [54]...55
第四章 實驗結果與討論...67
4.1 切削監測系統...68
4.1.1 切削訊號 ...68
4.1.2 切削相關數據...69
4.1.3 切削NC Code...70
4.2 防碰撞系統...71
4.2.1 刀具切削負載分析...72
4.2.2 刀具切削磨耗負載分析...77
4.2.3 防碰撞模型建立...82
4.2.4 刀具狀態監測介面...90
4.3 刀具磨耗預測系統...92
4.3.1 刀長設定器重複精度量測...95
4.3.2 磨耗分析 ...102
4.3.3 特徵值擷取...123
4.3.4 刀具磨耗預測結果...124
4.3.5 刀具磨耗預測結果驗證...127
第五章 結論與未來展望...138
5.1 結論...138
5.2 未來展望 ...139
參考文獻 ...140
Extended Abstract...148

[1]台灣區工具機暨零組件工業同業工會，2019，https://www.tmba.org.tw/message_show.php?cid=1448863855&pid=1444896759.
[2]N. I. Corporation., 2019, General Purpose Triaxial Accelerometers, http://sine.ni.com/nips/cds/view/p/lang/zht/nid/213178.
[3]K. I. Corporation., 2007, Acoustic Emission Senso, https://www.processcontrols.com/intertechnology/Kistler/pdfs/Accelerometer_Model_8152B.pdf.
[4]K. Group., 2019, Multi-Component Dynamometer up to 60 kN, https://www.kistler.com/en/product/type-9255c/.
[5]T. El-Wardany, D. Gao, M. Elbestawi, 1996, "Tool condition monitoring in drilling using vibration signature analysis", International Journal of Machine Tools and Manufacture, vol. 36, no. 6, pp. 687-711.
[6]R. Kuo, and P. Cohen, 1998, "Intelligent tool wear estimation system through artificial neural networks and fuzzy modeling", Artificial Intelligence in Engineering, vol. 12, no. 3, pp. 229-242.
[7]C. Chungchoo, and D. Saini, 2002, "On-line tool wear estimation in CNC turning operations using fuzzy neural network model", International Journal of Machine Tools and Manufacture, vol. 42, no. 1, pp. 29-40.
[8]J. Dong et al., 2006, "Bayesian-inference-based neural networks for tool wear estimation", The International Journal of Advanced Manufacturing Technology, vol. 30, no. 9-10, pp. 797-807.
[9]N. Ghosh et al., 2007, "Estimation of tool wear during CNC milling using neural network-based sensor fusion", Mechanical Systems and Signal Processing, vol. 21, no. 1, pp. 466-479.
[10]C. Aliustaoglu, H. M. Ertunc, and H. Ocak, 2009, "Tool wear condition monitoring using a sensor fusion model based on fuzzy inference system", Mechanical Systems and Signal Processing, vol. 23, no. 2, pp. 539-546.
[11]M. Malekian, S. S. Park, and M. B. Jun, 2009, "Tool wear monitoring of micro-milling operations", Journal of Materials Processing Technology, vol. 209, no. 10, pp. 4903-4914.
[12]康晶, 冯长建, 胡红英, 2009, "刀具磨损监测及破损模式的识别", 振動. 測試與診斷, vol. 29, no. 1, pp. 5-9.
[13]馬成傑，2007，”監督式學習類神經網路於銑削斷刀即時監控之研究”，中原大學工業與系統工程研究所碩士論文。
[14]黃耀賢，2010，”主軸振動與聲射訊號於微銑刀具磨耗監測之應用研究”，國立中興大學機械工程學系所碩士論文。
 
[15]M.-S. Noh and D. S. Hong, 2011, "Implementation of remote monitoring system for prediction of tool wear and failure using ART2", Journal of Central South University of Technology, vol. 18, no. 1, pp. 177-183.
[16]許育瑋，2012，”應用主軸振動與聲射訊號於鑽頭狀態偵測之研究”，國立中興大學機械工程學系所碩士論文。
[17]黃啟榮，2012，”切削路徑對應用主軸聲射與振動訊號之微銑削刀具磨耗偵測系統之影響分析”，國立中興大學機械工程學系所碩士論文。
[18]M. Rizal et al., 2013, "Online tool wear prediction system in the turning process using an adaptive neuro-fuzzy inference system", Applied Soft Computing, vol. 13, no. 4, pp. 1960-1968.
[19]林伯恂，2014，”切削條件改變下之刀具狀態監測”，國立臺灣大學機械工程學研究所碩士論文。
[20]姚皇任，2014，”以多感測器及類神經網路監測車削中刀具磨耗程度”，國立高雄第一科技大學機械與自動化工程研究所碩士論文。
[21]M. Ahmad et al., 2015, "Development of tool wear machining monitoring using novel statistical analysis method, I-kaz™", Procedia Engineering, vol. 101, pp. 355-362.
 
[22]R. Blödorn et al., 2015, "Study of the drilling process used in the Hole-Drilling Method through thrust force measurement and tool wear analysis", 23rd ABCM International Congress of Mechanical Engineering, pp. 1-8.
[23]S. Karam et al., 2016, "Online prediction of cutting tool life in turning via cognitive decision making", Procedia CIRP, vol. 41, pp. 927-932.
[24]朱永猛, 吴军, 吴超勇, 苏永衡, 2016, "基于自适应神经模糊推理系统的刀具磨损监测", 兵器装备工程学报, vol. 9, pp. 115-118.
[25]黃門，2016，”以電流及振動訊號監測微銑削刀具磨耗之研究”，國立臺灣大學機械工程學研究所碩士論文。
[26]R. Corne et al., 2017, "Study of spindle power data with neural network for predicting real-time tool wear/breakage during inconel drilling", Journal of Manufacturing Systems, vol. 43, pp. 287-295.
[27]J. Zhang et al., 2017, "Particle learning in online tool wear diagnosis and prognosis", Journal of Manufacturing Processes, vol. 28, pp. 457-463.
[28]陈永鹏, 曹华军, 杨潇, 2017, "高速干切滚齿工艺滚刀切削刃载荷分布特性研究", 机械工程学报, vol. 53, no. 15, pp. 181-187.
[29]梁宸瑜，2017，”SOM類神經網路濾波器於鎳基材料切削刀具磨耗監測系統之設計應用”，國立中興大學機械工程學系所碩士論文。.
[30]B. Jose et al., 2018, "Online Monitoring of Tool Wear and Surface Roughness by using Acoustic and Force Sensors", Materials Today: Proceedings, vol. 5, no. 2, pp. 8299-8306.
[31]鄭力維，2018，”工具機等切削力控制與刀具磨耗關係之探討”，國立臺灣大學機械工程學研究所碩士論文。
[32]林鴻熊，2009，”端銑刀銑削SKD11硬鋼之刀具磨耗研究”，華梵大學機電工程學系博碩專班碩士論文。
[33]T.-S. Lan, 2010, "Tool wear optimization for general CNC turning using fuzzy deduction", Engineering, vol. 2, no. 12, p. 1019.
[34]楊忠祥，2012，”S50C 中碳鋼精銑削刀具壽命之研究”，國立中興大學機械工程學系所碩士論文。
[35]O. Geramifard et al., 2014, "Multimodal hidden Markov model-based approach for tool wear monitoring", IEEE Transactions on Industrial Electronics, vol. 61, no. 6, pp. 2900-2911.
[36]潘建新, 潘祎, 2014, "刀具磨损与切削用量关联度应用与试验研究", 机械研究与应用, vol. 27, no. 5, pp. 102-104.
[37]C. Drouillet et al., 2016, "Tool life predictions in milling using spindle power with the neural network technique," Journal of Manufacturing Processes, vol. 22, pp. 161-168.
 
[38]K. Patra et al., 2017, "Artificial neural network based tool condition monitoring in micro mechanical peck drilling using thrust force signals", Precision Engineering, vol. 48, pp. 279-291.
[39]T. Mikołajczyk et al., 2018, "Predicting tool life in turning operations using neural networks and image processing", Mechanical systems and signal processing, vol. 104, pp. 503-513.
[40]高健瑋，2018，”車削製程刀具磨耗智慧預測系統之研發”，國立高雄應用科技大學機械工程系碩士論文。
[41]國立虎尾科技大學-智能機械與智慧製造研究中心，2019，Smart Machine Engine產品型錄。
[42]P. Technology, 2019, OPC UA Gateway to the IIoT, https://www.prosoft-technology.com/Landing-Pages/OPC-UA
[43]程泰機械股份有限公司，2019，GLS-1500 系列規格，http://www.goodwaycnc.com/exhtml_goodway/goodway_tw/turning/horizontal/gls-1500/.
[44]沅銘企業，2019，外徑車刀架，https://www.yuanminn.com.tw/zh/product_item/105.html.
[45]世鋁網，2019，鋁合金車削過程中積屑瘤的產生，https://big5news.cnal.com/2009/07-03/124662234162376.shtml.
 
[46]Renishaw，2019，HP 臂和RP3刀具設定測頭，https://www.renishaw.com.tw/tw/hp-arms-and-rp3-tool-settingprobe-6098
[47]T. M. Elektronik., 2019, RAYEX ELECTRONICS LEG-5F, https://www.tme.eu/en/details/leg-5f/miniature-electromagnetic-relays/rayex-electronics/.
[48]Plone基金會，2019，X ∕ Z 兩軸的定義及座標定義，http://support.intek.com.tw
[49]維基百科，2019，機器學習，https://zh.wikipedia.org/wiki/%E6%9C%BA%E5%99%A8%E5%AD%A6%E4%B9%A0.
[50]維基百科，2019，DeepMind，https://zh.wikipedia.org/wiki/DeepMind.
[51]GCP專家部落格，2019，人工智慧、機器學習、深度學習是什麼，https://blog.gcp.expert/ml-1-ai-ml-deep-learning-intro/.
[52]富捷IT培訓，2018，機器學習的4種模型，https://blogs.geego.com/skillsinfo/%E6%A9%9F%E5%99%A8%E5%AD%B8%E7%BF%92%E7%9A%844%E7%A8%AE%E6%A8%A1%E5%9E%8B%EF%BC%9A%E7%9B%A3%E7%9D%A3%E3%80%81%E9%9D%9E%E7%9B%A3%E7%9D%A3%E3%80%81%E5%8D%8A%E7%9B%A3%E7%9D%A3%E5%8F%8A%E5%A2%9E%E5%BC%B7/.
[53]維基百科，2019，監督式學習，https://zh.wikipedia.org/wiki/%E7%9B%A3%E7%9D%A3%E5%BC%8F%E5%AD%B8%E7%BF%92.
[54]林俊良，2016，智慧型控制：分析與設計(第三版)，全華圖書。
[55]T. Huang，2019，機器學習-神經網路(多層感知機詳細推導，https://medium.com/@chih.sheng.huang821/%E6%A9%9F%E5%99%A8%E5%AD%B8%E7%BF%92%E7%A5%9E%E7%B6%93%E7%B6%B2%E8%B7%AF%E5%A4%9A%E5%B1%A4%E6%84%9F%E7%9F%A5%E6%A9%9FmultilayerperceptronmlpE5%90%AB%E8%A9%B3%E7%B4%B0%E6%8E%A8%E5%B0%8Eee4f3d5d1b41.
[56]M. Opengate，2019，使用激勵函數的目的、如何選擇激勵函數，https://mropengate.blogspot.com/2017/02/deep-learning-role-of-activation.html.