臺灣博碩士論文加值系統

線切割放電加工應用中，不可避免會遇到工件厚度改變之需求，本文目的在於建立放電加工於變厚度製程中，可兼顧加工效率、穩定性以及槽寬一致性的深度學習智慧型控制系統。過去線切割放電加工於變厚度的控制，多聚焦在加工效率與穩定性的提升，對於同時考量槽寬不一現象影響工件尺寸精度的問題之研究則是較為欠缺的，從切割截面上可以得知加工後真平度不佳，且將可能導致未來工件配合上契合度不良等問題。文中利用深度學習(Deep Learning)藉著把資料透過多個非線性隱藏層的轉換，自動萃取出足以代表資料特性的特徵，亦即從多樣且大量的放電訊號中找出與工件厚度高度相關的隱性特徵函數且具有良好的強健性，不易受到雜訊或是少部份極值的影響，因此可準確得知工件厚度。
第二階段為加工參數的設定，在過去多依賴經驗豐富的操作人員，因此再次利用深度學習的技術，使輸入資料的隱性特徵經過歸一化處理後，讓模型輸出猶如豐富經驗操作員的判定。如此的智慧型控制策略不只可以減少操作人員技術門檻，各厚度槽寬均一變動量小的目標，可以使第二刀精修時的材料移除率更為穩定，減少線痕的產生。
由實驗結果顯示，本文所建立的深度學習工件厚度線上估測系統，可以準確估測出工件厚度，且加工參數智慧型控制系統，其可根據不同的工件厚度調整放電休止時間與輔助放電休止時間，使放電頻率被控制在期望值之內，在工件厚度由薄變厚時，可有效提升加工效率；在工件厚度由厚變薄時，可以改善進給速度過快與放電集中所導致斷線的現象。另一方面，本研究之策略有效兼顧了槽寬一致性，亦即加工不同工件厚度時，其各階級厚度之槽寬平均值與變動量可有效維持在穩定的數值內。

The purpose of this study is to establish a deep learning intelligent control system that can combine processing efficiency, stability and kerf consistency in variable thickness processing. In the past, the wire electrical discharge machining Control for variable thickness workpiece, and focus on improving processing efficiency and stability. The research on the problem that the kerf difference affects the accuracy of the variable thickness workpiece is relatively lacking. The true straightness is not good after processing, and it may lead to error fitting of future workpieces. Deep Learning is used to automatically extract features that are representative of data characteristics by transforming data through multiple nonlinear hidden layers, that is, from a variety of discharge signals that are highly correlated with the thickness of the workpiece. The recessive feature function has strong robustness and is not sensitive to noise or some of extreme values, so the thickness of the workpiece can be accurately known.
The second stage is the setting of processing parameters. In the past, it relied on experienced operators. Therefore, the deep learning technique was used again to normalize the hidden features of the input data, so that the model output is like a experience operator to determination what the parameters should be setting. Intelligent control strategy not only reduce the operator''s technical threshold, but also achieve a goal of uniform material removal rate during the trim cutting and reduce the occurrence of line marks.
The experimental results show that the deep learning workpiece thickness online estimation system established in this study which can accurately estimate the thickness of the workpiece, and the processing parameter intelligent control system can adjust the discharge TOFF and TAOFF according to different workpiece thicknesses. The discharge frequency is controlled within the desired value, and the processing efficiency can be effectively improved when the thickness of the workpiece is thin to thick; when the thickness of the workpiece is thick to thin, the feed rate will be accelerated and the wire is broken by discharge concentration phenomenon. On the other hand, the strategy of this study effectively balances the kerf uniformity, that is, the average value and variation of the kerf of each class thickness can be effectively maintained within a stable value when processing different workpiece thicknesses.

誌謝 I
摘要 II
Abstract III
目錄 V
圖目錄 VIII
表目錄 XI
符號說明 XII
第一章 緒論 1
1.1 研究背景與動機 1
1.2 文獻回顧 3
1.3 研究目的與研究方法 8
1.4 本文結構 9
第二章 放電加工相關技術與原理 10
2.1 放電現象簡述與材料移除機制 10
2.2 放電電弧的發展過程 12
2.3 放電火花結構 13
2.4 放電迴路的種類及原理 15
2.5 放電參數與放電波形 17
2.6 工件電解現象與AC Power電源 19
2.6.1 DC Power之電解現象與對策 19
2.6.2 AC Power之正負向點火邏輯 21
2.7 放電加工機進給控制系統 22
第三章 深度學習之理論 24
3.1 深度學習的發展歷史 25
3.2 深度學習的基石演算法 25
3.2.1 感知器(Perceptron) 25
3.2.2 多層感知器(Multilayer Perceptron) 27
3.2.3 深度學習模型的分類 28
3.2.4 深度學習網路的運作原理 31
3.2.5 深度學習網路的特性 32
3.3 深度學習神經網路 33
3.3.1 倒傳遞類神經網路簡介 33
3.3.2 倒傳遞類神經網路架構 33
3.3.3 倒傳遞類神經網路學習演算法 34
3.3.4 激勵函數(Active Function) 37
3.3.5 損失函數(Loss Function) 40
3.3.6 最佳化演算法(Optimization Algorithm) 41
3.3.7 過度擬合(Overfitting) 43
3.3.7 Dropout 44
第四章 實驗設備與實驗規劃 46
4.1 實驗設備與材料 47
4.2 放電訊號之擷取與處理 55
4.2.1 光耦合隔離電路 56
4.2.2 數位訊號整形電路 56
4.2.3 差動傳輸原理 57
4.3 放電即時監控系統的驗證 60
4.4 線上網路傳輸控制系統 61
第五章 實驗結果與討論 63
5.1 工件厚度線上估測模型建立 63
5.2 加工參數智慧型控制模型建立 65
5.3 控制策略之架構圖 69
5.4 控制方法之分析與比較 70
5.4.1 傳統伺服控制於工件厚度增加之現象探討 71
5.4.2 智慧型控制於工件厚度增加之現象探討 74
5.4.3 傳統伺服控制於工件厚度減少之現象探討 77
5.4.4 智慧型控制於工件厚度減少之現象探討 80
5.5 控制結果討論 83
5.5.1 加工效率 83
5.5.2 加工穩定度與槽寬 86
第六章 結論與建議 94
6.1 變厚度智慧型控制之研究結論 94
6.2 未來展望 95
參考文獻 96

[1]K. K. Jangra, V. Kumar, and V. Kumar, "An Experimental and Comparative Study on Rough and Trim Cutting Operation in WEDM of Hard to Machine Materials," Procedia Materials Science, vol. 5, pp. 1603-1612, 2014.
[2]K. K. Jangra, "An experimental study for multi-pass cutting operation in wire electrical discharge machining of WC-5.3% Co composite," The International Journal of Advanced Manufacturing Technology, vol. 76, no. 5, pp. 971-982, 2015.
[3]K. P. Rajurkar, W. M. Wang, and J. A. McGeough, "WEDM Identification and Adaptive Control for Variable-Height Components," CIRP Annals, vol. 43, no. 1, pp. 199-202, 1994.
[4]K. P. Rajurkar, W. M. Wang, and W. S. Zhao, "WEDM-Adaptive Control with a Multiple Input Model for Indentification of Workpiece Height," CIRP Annals, vol. 46, no. 1, pp. 147-150, 1997.
[5]L. Zhidong, C. Haoran, P. Huijun, Q. Mingbo, and T. Zongjun, "Automatic control of WEDM servo for silicon processing using current pulse probability detection," The International Journal of Advanced Manufacturing Technology, vol. 76, no. 1, pp. 367-374, 2015.
[6]J. Wang, F. Han, and F. Zhao, "Improvement of EDM efficiency with a new adaptive control strategy," The International Journal of Advanced Manufacturing Technology, vol. 62, no. 9, pp. 1025-1040, 2012.
[7]K. T. Hoang and S. H. Yang, "Kerf analysis and control in dry micro-wire electrical discharge machining," The International Journal of Advanced Manufacturing Technology, vol. 78, no. 9, pp. 1803-1812, 2015.
[8]K. T. Hoang and S.-H. Yang, "A new approach for Micro-WEDM control based on Real-Time estimation of material removal rate," International Journal of Precision Engineering and Manufacturing, vol. 16, no. 2, pp. 241-246, 2015.
[9]I. Maher and A. A. D. Sarhan, "Proposing a new performance index to identify the effect of spark energy and pulse frequency simultaneously to achieve high machining performance in WEDM," The International Journal of Advanced Manufacturing Technology, vol. 91, no. 1, pp. 433-443, 2017.
[10]H. Bisaria and P. Shandilya, "Experimental investigation on wire electric discharge machining (WEDM) of Nimonic C-263 superalloy," Materials and Manufacturing Processes, vol. 34, no. 1, pp. 83-92, 2019.
[11]李其政, 類神經網路應用於線切割放電加工厚度變化工件之研究, 國立台灣大學機械工程學研究所碩士論文, 2004.
[12]S.-C. Dou, X.-C. Xi, and W.-S. Zhao, "Workpiece height estimation in wire electrical discharge machining by using support vector regression," Proceedings of the Institution of Mechanical Engineers, Part B: Journal of Engineering Manufacture, vol. 227, no. 4, pp. 565-577, 2013.
[13]G. Huang, W. Xia, L. Qin, and W. Zhao, "Online Workpiece Height Estimation for Reciprocated Traveling Wire EDM Based on Support Vector Machine," Procedia CIRP, vol. 68, pp. 126-131, 2018.
[14]林奕圻, 模糊控制應用於線切割放電加工厚度變化工件之研究, 國立台灣大學機械工程研究所碩士論文, 2005.
[15]W. M. Lee and Y. S. Liao, "Adaptive control of the WEDM process using a self-tuning fuzzy logic algorithm with grey prediction," The International Journal of Advanced Manufacturing Technology, vol. 34, no. 5, pp. 527-537, 2007.
[16]Y. S. Liao, T. J. Chuang, and Y. P. Yu, "On-line Workpiece Height Estimation and its Application in Servo Feed Control of WEDM Process," Procedia CIRP, vol. 6, pp. 226-231, 2013.
[17]W.-M. Lee and Y.-S. Liao, Self-tuning fuzzy control with a grey prediction for wire rupture prevention in WEDM. 2003.
[18]M. T. Yan, Y. S. Liao, and C. C. Chang, "On-line Estimation of Workpiece Height by Using Neural Networks and Hierarchical Adaptive Control of WEDM," The International Journal of Advanced Manufacturing Technology, journal article vol. 18, no. 12, pp. 884-891, 2001.
[19]J. Sanchez, A. Conde, A. Arriandiaga, J. Wang, and S. Plaza, "Unexpected Event Prediction in Wire Electrical Discharge Machining Using Deep Learning Techniques," Materials, vol. 11, no. 7, p. 1100, 2018.
[20]李强, 电火花线切割加工工件厚度在线识别技术研究, 哈尔滨工业大学机电工程学院黑龙江哈尔滨, 2016.
[21]陈良 and 张红玉, 多输入工件厚度辨识模型的LS-WEDM自适应控制系统, 机械设计与制造, no. 1, pp. 101-103, 2017.
[22]李國祥, WEDM細線加工之放電能量與電流波形探討, 國立台北科技大學電機工程學研究所碩士論文, 2004.
[23]林季穎, 基於深度學習之人臉特徵辨識與應用, 國立交通大學光電工程學研究所碩士論文, 2018.
[24]任博仲, 基於深度學習之電影預告片分類, 中學大學資訊工程學研究所博士論文, 2019.
[25]鄭宇廷, 利用深度學習探勘地物結構及異常體, 國立台北科技大學土木工程學研究所碩士論文 2018.
[26]劉人傑, 基於深度學習之人聲辨識探討資料集組成與測試準確率之關聯性, 國立台北科技大學資訊工程學研究所碩士論文, 2018.
[27]A. Muniappan, R. Prasanna, T. Shaafi, V. Jayakumar, and M. Ajithkumar, "Investigation of wire EDM control variables on kerf width for Al6063/SiC/Al2O3 composite by response surface methodology," IOP Conference Series: Materials Science and Engineering, vol. 402, p. 012152, 2018.
[28]V. Parashar, A. Sidiqui, J. L. Bhagoria, and Y. M. Puri, Kerfs width analysis for wire cut electro discharge machining of SS 304L using design of experiments., pp. 369-373. 2010
[29]D. V. S. S. S. V. Prasad and A. G. Krishna, "Empirical modeling and optimization of kerf and wire wear ratio in wire electrical discharge machining," The International Journal of Advanced Manufacturing Technology, vol. 77, no. 1, pp. 427-441, 2015.
[30]廖哲暐, 線切割放電加工機鼓形量之探討與抑制, 國立台灣大學機械工程學研究所碩士論文, 2010.