臺灣博碩士論文加值系統

在智能化與無人化技術高速發展的情況下，世界各國在精進工業技術上都以能夠達到智慧工廠與相關技術為核心發展目標，近年來我國工業在科技發展下在工業4.0上已有大幅的突破，隨著國內少子化的情形越來越嚴重，導致勞動人口降低大幅增加企業人力成本，現今加工廠都需要有經驗的技師操作工具機，而人為操作可能會因為受到各種因素影響而導致失誤，在智能化工廠之技術發展下可以有效解決人力短缺之與人為操作的風險。在講求高度精密加工的情況下，使得磨床的相關技術也顯得越來越重要，磨削加工應用的增長幅度遠超過其他傳統加工方法，並深知提升磨床加工效率與降低操作難度是相當重要的，在國內外磨削加工中，以傳統往復方式研磨工件通常都會存在加工效率低落的問題，由於在實際加工情況下工件胚料的外形千變萬化，以傳統方式路徑研磨工件時無法隨著工件外型變化加工路徑，導致以傳統矩形路徑研磨工件的方式容易導致大部分的磨削路徑是屬於空跑狀態，因此如何讓機台自行判斷優化之加工路徑進行磨削為新興課題，故本研究針對CNC磨床提出以機器視覺的方式對平面磨削加工路徑進行優化，以達到提升平面磨削加工效率並簡化操作方式，使用機器視覺替代部分人為判斷。本研究論文主要可以分為四部分探討，第一部分為將工業相機架設於主軸側邊隨著工件高度的設定移動主軸使拍照高度穩定，並架設白色LED條狀光源於機台Z軸方向，利用電磁鐵反光的特性，使床台產生近似背光照明以凸顯工件胚料輪廓，並以工業相機進行影像擷取；第二部分為使用EmguCV影像處理函式庫將擷取到之工件輪廓影像進行濾波、Otsu二值化與邊緣偵測計算工件邊緣像素資訊；第三部分為利用所擷取的邊緣資訊與本研究平面磨削加工路徑生成邏輯，依據設定之加工參數篩選出所需要的加工路徑點；第四部分為利用建立在控制器中平面研磨Macro動作模板，依照平面磨削加工邏輯套用影像系統所篩選的加工路徑點，使其能夠自動依照工件輪廓進行加工，達成機器視覺整合與CNC磨床當中。實驗結果顯示，在光源穩定之情況下利用工業相機擷取工件輪廓之尺寸誤差在3.5%內；以本研究所生成之路徑與傳統平面磨削路徑相互比較，依據所加工的工件胚料輪廓不同，能夠縮減總路徑長度49.26%至7.5%，本研究與傳統平面磨削方式相比能夠提升加工效率，並且達到簡化整體操作之目的。

With the rapid development of intelligent and unmanned technologies, all countries in the world have been able to achieve smart factories and related technologies as their core development goals in improving industrial technology. The declining birthrate has become more and more serious, leading to a loss in the labor force and a significant increase in the labor cost of enterprises. Nowadays, manufacturers require experienced technicians to operate machine tools. People may make mistakes due to various factors. However, there are some problems with processing efficiency in the traditional reciprocating grinding process. Due to the ever-changing contours of the workpiece, the traditional grinding processing path cannot adjust with the shape of the workpiece, so how to let the machine adjust the processing path by workpiece contour is an emerging topic. This study proposed a method to optimize the surface grinding processing path using machine vision, which can improve the overall processing efficiency and simplify the operation method. This research can be divided into four parts to discuss. The first part is setting the industrial camera on the side of the spindle and moving the spindle with the height of the workpiece to stabilize the working distance, and setting up the white LED strip light source in the Z-axis to make the contour of the workpiece stand out, and use the industrial camera to capture the image; the second part is using EmguCV image processing library to calculate the contour information of the workpiece; the third part is using the contour information to generate optimized grinding processing path according to the processing parameters and logic of surface grinding; the fourth part is to combine edge points and grinding logic to generate optimized processing paths by self-design MACRO in the controller. The results of the experiments show that the workpiece size error of the vision system is within 3.5%; the optimized path generated by this system can reduce the total tool path from 49.26% to 7.5%, which can improve machining efficiencies and achieve the purpose of simplifying operation.

摘要............i
Abstract............ii
誌謝............iii
目錄............iv
表目錄............vi
圖目錄............vii
符號說明............x
第 一 章 緒論............1
1.1 研究背景與動機............1
1.2 研究目的............2
1.3 論文架構............2
第 二 章 文獻回顧 ............3
2.1 工具機與機器視覺之應用相關研究............3
2.2 國內外相關研究............8
第 三 章 研究架構與方法............12
3.1 研究架構與流程............12
3.2 硬體及系統架構............13
3.2.1 影像擷取硬體設備............13
3.2.2 實驗機台介紹............14
3.2.3 數位影像照明之選用與架設方式............14
3.3 相機成像與校正原理............17
3.3.1 相機成像原理............17
3.3.2 世界坐標與相機成像坐標轉換............18
3.4 數位影像處理演算法............19
3.4.1 數位影像組成............20
3.4.2 色彩空間轉換與影像灰階............20
3.4.3 影像空間濾波............22
3.4.4 直方圖均衡化............23
3.4.5 影像二值化............25
3.4.6 型態學影像處理............27
3.4.7 邊緣偵測............29
3.5 平面磨削加工路徑生成............30
3.6 CNC磨床機台通訊與資料傳遞............32
3.6.1 新代控制器通訊與連線............32
3.6.2 影像處理系統架構............34
3.6.3 新代MACRO程式設計與撰寫............35
第 四 章 實驗結果............39
4.1 工件影像擷取系統............40
4.1.2 相機畸變校正............41
4.1.3 智能化影像辦識主程式............42
4.1.4 環境光譜分析............43
4.1.5 工件位置對於影像辨識精度影響實驗............44
4.1.6 工件尺寸對於影像辨識精度實驗............46
4.2 自動影像路徑生成系統............46
4.2.1 自動影像路徑生成系統操作介面............47
4.2.2 影像路徑生成實驗-倒T型............48
4.2.3 影像路徑生成實驗-圓型............49
4.2.4 影像路徑生成實驗-金字塔型............50
4.2.5 影像直方圖與路徑比較............52
4.2.6 預估加工效率比較-倒T型............54
4.2.7 預估加工時間比較-圓型............55
4.2.8 預估加工時間比較-金字塔型............55
4.2.9 加工路徑效率實驗-倒T型............56
4.2.10 加工路徑效率實驗-圓型............57
4.2.11 加工路徑效率實驗-金字塔型............58
第 五 章 結論與未來展望............60
參考文獻............61
附錄一 平面研磨加工動作模板............63
附錄二 G200正向研磨............65
附錄二 G200正向研磨............66
附錄三 G201反向研磨............67
附錄四 G500定位拍照............68
附錄五 實驗資料............69
Extended Abstract............75

[1]R. S. Woodbury, 1959. “History of the Grinding Machine. Technology Press”, Cambridge, MA.
[2]C. Andrew, T. D. Howes, T. R. Pearce, 1985. “Creep Feed Grinding”. Holt, Rinehart &Winston Ltd, Eastbourne.
[3]X. Chen, T. T. Öpöz, “Effect of different parameters on grinding efficiency and its monitoring by acoustic emission”, Production & Manufacturing Research, Pages 190-208, 2016.
[4]J.K. Che, M.M. Ratnam, “Real-time monitoring of workpiece diameter during turning by vision method”, Measurement, Pages. 369-377, 2018.
[5]L. M. Xu, F. Fan, Z. Zhang, Y. Chen, D. J. Hu, L. Shi, “Methodology and implementation of a vision-oriented open CNC system for profile grinding”, The International Journal of Advanced Manufacturing Technology, Volumes 100, Pages 9-12, February 2019.
[6]L. M. Xu, M. Niu, D. Zhao, N. Xing, F. Fan, “Methodology for the immediate detection and treatment of wheel wear in contour grinding”, Precision Engineering, Volumes 60, Pages 405-412, November 2019.
[7]H. K. Abdul-Ameer, G. A. Al-Kindi, H. Zughaer, “Towards Computer Vision Feedback for Enhanced CNC Machining”, IEEE 3rd International Conference on Communication Software and Networks, May 2011.
[8]A. A. Khleif, A. A. Shnawa, “Vision system aided 3D object reconstruction and machining using CNC milling machine”, International Conference on Advance of Sustainable Engineering and its Application (ICASEA), March 2018.
[9]N. Wan, C. Xia, H. Zhao, S. Zhang, “5-axis grinding path generation for free-form surface based on plane instantaneous grinding engagements”, International Journal of Advanced Manufacturing Technology, Volumes 103, Pages 1861-1877, April 2019.
[10]J. Xie, R. M. Zhou, J. Xu, Y. G. Zhong, “Form-truing error compensation of diamond grinding wheel in CNC envelope grinding of free-form surface”, The International Journal of Advanced Manufacturing Technology, Volume 48, Pages 905-912, October 2009.
[11]J. Xie, J. H. Zheng, R. M. Zhou, B. Lin, “Dispersed grinding wheel profiles for accurate freeform surfaces”, International Journal of MachineTools & Manufacture, Volume 51, Pages 536-542, March 2011.
[12]M. Dhanda, S. S. Pande, “Adaptive Tool Path Planning Strategy for Freeform Surface Machining using Point Cloud”, Computer-Aided Design and Applications, Volume 16, Pages 289-307, January 2019.
[13]R. Ahmad, P. Plapper, “Generation of safe tool-path for 2.5D milling/drilling machine-tool using 3D ToF sensor”, CIRP Journal of Manufacturing Science and Technology, Volume 10, Pages 84-91, August 2015.
[14]R. C. Gonzalez, R. E. Woods, “Digital Image Processing”, Pearson FT Press, 2017.
[15]江孟桓, “智能化工具機即時加工檢測與伺服系統調機”，國立虎尾科技大學，碩士學位論文, 2015。
[16]李信良, “自動化斜進式外圓磨床之工業4.0技術應用”, 國立虎尾科技大學, 碩士學位論文, 2016。
[17]許嘉洧, “基於格雷碼結構光輔助工具機之工件自動定位”, 國立虎尾科技大學, 碩士學位論文, 2018。
[18]EmguCV, “http://www.emgu.com/wiki/index.php/Main_Page”。
[19]株式会社岡本工作機械製作所官網, “http://www.okamotoen.com/machines”。
[20]S. Suzuki, “Topological structural analysis of digitized binary images by border following”, Computer Vision, Graphics, and Image Processing, Volume 30, Pages 32-46, 1985.
[21]J. G. Gohn, A. N, “Illumination Compensated images for surface roughness evaluation using machine vision in grinding process”, Procedia Manufacturing, Volume 34, Pages 969-977, 2019.
[22]R Manish, A. Venkatesh, S. Denis Ashok, “Machine Vision Based Image Processing Techniques for Surface Finish and Defect Inspection in a Grinding Process”, Materialstoday:Proceedings, Volume 5, Pages 12792-12802, 2018.
[23]B. A. Abu-Nabah, A. O. ElSoussi, A. K. Al Alami, “Virtual laser vision sensor environment assessment for surface profilingapplications”, Measurement, Volume 113, Pages 148-160, January 2018.
[24]K. Wegener, F. Bleicher, P. Krajnik, H. W. Hoffmeister, C. Brecher, “Recent developments in grinding machines”, CIRP Annals - Manufacturing Technology, Volume 66, Pages 779-802, 2017.
[25]莊汶諺, “CNC工具機之視覺輔助定位系統研發”, 國立清華大學, 碩士學位論文, 2012。