(3.236.118.225) 您好!臺灣時間:2021/05/16 09:38
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果

詳目顯示:::

: 
twitterline
研究生:張烜睿
研究生(外文):Hsuan-Jui Chang
論文名稱:智能變壓器插件系統
論文名稱(外文):An Intelligent Robotic Transformer Insertion System
指導教授:林顯易
指導教授(外文):Hsien-I Lin
口試委員:林顯易黃中明宋開泰
口試委員(外文):Hsien-I LinHsien-I LinHsien-I Lin
口試日期:2016-07-29
學位類別:碩士
校院名稱:國立臺北科技大學
系所名稱:自動化科技研究所
學門:工程學門
學類:機械工程學類
論文種類:學術論文
論文出版年:2016
畢業學年度:104
語文別:中文
中文關鍵詞:腳位辨識、AOI、動作分類、機器學習
外文關鍵詞:Insertion、AOI、Machine learning
相關次數:
  • 被引用被引用:0
  • 點閱點閱:149
  • 評分評分:
  • 下載下載:24
  • 收藏至我的研究室書目清單書目收藏:0
在電子業界裡,電子元件的插件技術是極為重要的關鍵技術,過去時常運用大量的人力分工組裝,演變至今客製化的專用機自動化技術,若本研究中變壓器只考慮單一插件動作,會捨去許多可以插件但不符合此動作的零件;此外變壓器的六支針腳皆為人工焊接,所以有六個變異量存在,因此使用專用機的快速插件有很大的困難。本研究提出結合影像特徵與多種機器手臂插件動作的學習系統,改善目前無法使用專用機對對變壓器(異型件)的插件問題。本研究目的運用自由度較高的六軸機械手臂解決變壓器的插件,以及學習變壓器影像特徵與機器手臂插件動作關聯,來預測變壓器的插件動作。本研究的方法分成影像分析層(vision layer)、插件動作層(motion layer)、影像動作關係決策層(decision layer)。影像分析層運用影像的前處理,進行變壓器的瑕疵檢測與擷取變壓器腳位的影像特徵,以及建立後續學習機器手臂插件動作的重要參數;動作層收集符合變壓器插件的動作姿態,並運用Fuzzy C-means收斂插件動作的種類,以建立決策層輸出的目標標記。決策層使用階層的(Support Vector Machines, SVM)建立影像特徵與插件動作之間的分類器,用影像特徵來預測插件動作。本研究使用300顆變壓器當訓練樣本,200顆變壓器當測試樣本,實驗裡運用不同的影像特徵限制條件,分別比較SVM、貝氏網路與類神經網路對測試樣本的準確度,本研究所使用三層結構方法,可以達到87%的準確度,ROC曲線的AUC達0.88,較貝氏網路與類神經網路的準確度高。
In electronics industry, the automatic insertion technology of electronic component plays an important role. Traditionally, it is necessary to identify the geometrical shape of insertion objects to find a proper insertion pose. Even though some of insertion tasks can be done by an 4-DOF robot arm (SCARA), this SCARA robot has more workspace constraints than an six-DOF robot. In this research, we aim to solve the problem of automatic insertion for transformers.Transformers are difficult to insert because there are 6 pins in a transformer to be inserted at the same time. Since a SCARA robot is not suitable in this problem, we use a 6-DOF robot to validate the experiments. In this research, we propose a three-layer method including vision, motion, and decision layers. The vision layer is to extract important features of transformers for the decision layer. The motion layer is constructed by using Fuzzy C-means (FCM) to find representative insertion patterns. The decision layer based on Support Vector Machine (SVM) is used to predict the insertion pose for transformers. By training a great number of transformers, the hierarchical SVMs learn the relationship of vision features of transformer pins and the corresponding insertion poses. The result showed that the accuracy rate of the proposed method on five hundred testing transformers is up to 87%.
摘 要 i
ABSTRACT ii
誌 謝 iv
目 錄 v
表目錄 vii
圖目錄 viii
第一章 緒論 1
1.1 背景與動機 1
1.2 研究目的 2
1.3 研究大綱 3
1.4 研究貢獻 4
第二章 文獻探討 5
2.1 機器人插件 6
2.1.1 基於力回饋 6
2.1.2 插件收尋策略 7
2.1.3 基於視覺 8
2.1.4 基於力回饋與視覺 8
2.2 自動化光學檢測(Automatic Optical Inspection) 9
2.2.1 圖形識別 9
2.2.1.1 模板匹配 9
2.2.1.2 特徵比對 10
2.2.1.3 圖形結構法 10
2.2.2 圖形分割 11
2.2.3 影像特徵擷取方法 11
第三章 影像處理基礎方法與動作分群 13
3.1 光源 13
3.1.1 燈源總類 13
3.1.2 打光技術 14
3.2 色彩模型(Color Space) 18
3.2.1 RGB色彩空間 18
3.2.2 YCbCr色彩空間 19
3.3 影像二值化 20
3.4 形態學(Morphological) 23
3.4.1 數學定義 23
3.4.2 閉合(Closing) 23
3.4.3 斷開(Opening) 24
3.5 連通區域標記(Connected-component labeling) 25
3.6 資料維度減少 27
3.5.1 主成分分析(Principal Component Analysis) 27
3.5.2 LLE(Locally Linear Embedding) 29
3.7 動作分群 32
3.8 支持向量機(Support Vector Machine) 35
3.8.1 SVM原理介紹 36
3.8.2 非線性SVM 37
第四章 硬體架構 39
4.1 硬體系統架構 39
4.2 機器手臂 41
4.3 電動夾爪 42
4.4 工業相機 42
4.5 選用之光源技術 44
第五章 插件系統 46
5.1 系統概述 46
5.2 針腳影像處理 47
5.3 插件動作分群 56
5.3.1 人工收尋 57
5.3.2 FCM收尋 58
5.3.3 FCM(三維與權重)收尋 59
5.4 SVM 規劃 62
5.4.1 資料前處理 62
5.4.2 SVM分類設計 64
第六章 實驗 68
6.1 實驗一 68
6.2 實驗二 73
6.3 實驗三 75
第七章 結論與未來展望 78
參考文獻 80
[1]M. T. Mason, "Compliance and force control for computer controlled manipulators," IEEE Trans. System, Man and Cybernetics, vol. SMC-11, No.6, pp.418-432, 1981.
[2]M. H. Raibert and J. J. Craig, "Hybrid force/position control of manipulators," Amer. Soc. Mech. Eng. Winter Annu. Meeting, 1980.
[3]S. T. Lin and A. K. Huang, "Hierarchical fuzzy force control for industrial robot," IEEE Trans. Ind. Electron., vol. 45, no. 4, pp. 646-653, 1998.
[4]N. Hogan, "Impedance control: An approach to manipulation: Parts I-III," J. Dynamic Syst., Measure. Control, vol. 107, no. 1, pp. 1-24, 1985.
[5]V. Gullapalli, R. A. Grupen, and A. G. Barto, "Learning reactive admittance control," in Proc. IEEE International Conf. Robotics and Automation, 1992., pp. 1475-1481.
[6]M. D. Majors and R. J. Richards, “A neural-network-based flexible assembly controller,” in Fourth Int. Conf. Artificial Neural Networks, 1995, pp. 268-273.
[7]K. Sharma, V. Shirwalkar, and P. K. Pal, "Intelligent and environment-independent peg-in-hole search strategies," in IEEE Int. Conf. on Control, Automation, Robotics and Embedded Systems, 2013, pp. 1-6, December 16-18.
[8]J. Y. Kim, “Novel visual sensing systems for overcoming occlusion in robotic assembly,” Assembly Automation, vol. 25, no. 1, pp. 21-29, 2005.
[9]B. Tamadazte , S. Dembélé and N. L. Fort-Piat, "Robotic micromanipulation and microassembly using monoview and multiscale visual servoing," IEEE/ASME Trans. Mechatronics, vol. 16, no. 2, pp. 277-287, 2011.
[10]I. Lopez-Juarez, J. Corona-Castuera, M. Pena-Cabrera, K. Ordaz-Hernandez
"On the design of intelligent robotic agents for assembly," Information Sciences, 171, pp. 377-402, 2005.
[11]G. Morel, E. Malis, and S. Boudet, "Impedance based combination of visual and force control," in Proc. IEEE Int. Conf. Robotics and Automation, Leuven, Belgium, 1998, pp. 1743-1748.
[12]P. Chang, L. Chen, and C. Fan, "A case-based evolutionary model for defect classification of printed circuit board images," Journal of Intelligent Manufacturing, vol. 19, no. 2, pp. 203-214, 2008.
[13]C. J. Chen, S. H. Lai, W. H. Lee, C. Y. Lin, T. Ku, C. H. Chen, and Y. C. Chung,"Hausdorff distance based PCB inspection system with defect classification." Optics East 2005. International Society for Optics and Photonics, 2005.
[14]M. Crispin and V. Rankov, "Automated inspection of PCB components using a genetic algorithm template-matching approach," Int. J. Adv. Manuf. Technol., vol. 35, pp. 293-300, 2007.
[15]Y. Xie, Y. Ye, J. Zhang, L. Liu, L. Liu "A physics-based defects model and inspection algorithm for automatic visual inspection," Optics and Lasers in Engineering 52, pp. 218-223, 2014.
[16]
M. Moganti and F. Ercal, "Automatic PCB inspection systems," IEEE Potentials, vol. 14, no. 3, pp. 6-10, 1995.
[17]G. Acciani , G. Brunetti and G. Fornarelli, "Application of neural networks in optical inspection and classification of solder joints in surface mount technology", IEEE Trans. Ind. Informat., vol. 2, no. 3, pp. 200-209, 2006.
[18]K. W. Ko, H. S. Cho, J.H. Kim, and J. S. Kim, "Solder Joint Inspection using Neural Network and Fuzzy Rule-Based Classification," in Proc. IEEE/RSJ Int. Conf. Intelligent Robots and Systems, 1998, pp. 1565-1570.
[19]A. N. Belbachir, M. Lera, A. Fanni, and A. Montisci, "An automatic optical inspection system for the diagnosis of printed circuits based on neural networks," in Proc. Ind. Appl. Conf. , 2005, vol. 1, pp. 680-684.
[20]A. Giaquinto , G. Fornarelli , G. Brunetti, and G. Acciani, "A neuro-fuzzy method for the evaluation of soldering global quality index," IEEE Trans. Ind. Inf., pp. 56-66, 2009.
[21]M. Howarth and A. Lotfi, "Adaptive fuzzy control of solder paste printing: The identification of deposit defects," in Proc. Electronics Manufacturing Technology Symp., pp. 102-107, 1999.
[22]K. W. Kon and H. S. Cho, "Solder joints inspection using a neural network and fuzzy rule-based classification method," IEEE Trans. Electron. Packag. Manuf., vol. 23, no. 2, pp. 93-103, 2000.
[23]M. E. Zervakis, S. K. Goumas, and G. A. Rovithakis, "A Bayesian framework for multilead SMD post-placement quality inspection," IEEE Transactions Systems, Man, and Cybernetics, Part B, vol. 34, no. 1, pp. 440-453, 2004.
[24]A. Fanni, M. Lera, E. Marongiu, and A. Montisci, "Neural network diagnosis for visual inspection in printed circuit boards," in 12th International Workshop on Principles of Diagnosis, pp. 47-54, 2001.
[25]N. S. S. Mar, P. K. D. V. Yarlagadda, and C. Fookes, "Design and development of automatic visual inspection system for PCB manufacturing," Robot. Comput.-Integr. Manuf., vol. 27, no. 5, pp. 949-962, 2011.
[26]T. Abatzoglou and B. O’ Donnell, “Minimization by Coordinate Descent,” J. Optim. Theory Appl., vol. 36, pp. 163-174, 1982.
[27]J. R. Shewchuk, "An introduction to the conjugate gradient method without the agonizing pain," Pittsburgh, PA, USA, Tech. Rep., 1994.
[28]G. Hovland and B. J. McCarragher, "Frequency-Domain Force Measurements for Discrete Event Contact Recognition," IEEE International Conference on Robotics and Automation, 1996
[29]E. Cervera and A.P. del Pobil , "Eliminating Sensor Ambiguities Via Recurrent Neural Networks in Sensor-Based Learning," IEEE International Conference on Robotics and Automation, vol. 3, pp. 2174-2179, 1998.
[30]E. Cervera, A.P. Pobil, E. Marta and M.A. Serna, "A Sensor-Based Approach for Motion in Contact in Task Planning", Proceedings of IEEE/RSJ International Conference on Intelligent Robots and Systems, 1995, vol. 2, pp. 468-473.
[31]邱威堯, "打光技術",工業技術研究院 機器視覺部,2014。
[32]N. Otsu, "A threshold selection method from gray-level histogram," IEEE Trans. Syst. Man Cybern., vol. 9, pp. 62-66, 1979.
[33]Gonzalez, Woods, and Eddins, Digital Image Processing, 3rd edition, Prentice Hall, 2008.

[34]S. T. Roweis and L. K. Saul, "Nonlinear Dimensionality Reduction by Locally Linear Embedding," Science, vol. 290, Dec. 2000.
[35]C. Cortes and V. Vapnik, "Support-vector network," Machine Learning, vol. 20, pp. 273-297, 1995.
[36]B. Boser, I. Guyon, and V. N. Vapnik, "A training algorithm for optimal margin classifiers," in Proc. 5th Annu. Wkshp. Comput. Learning Theory, pp. 144-152, 1992.
[37]J. H. Min and Y.-C. Lee, "Bankruptcy prediction using support vector machine with optimal choice of kernel function parameters," Expert Syst. Appl., vol. 28, pp. 603-614, 2005.
[38]X. L. Xie and G. Beni,"A Validity Measure for Fuzzy Clustering" Fuzzy Models for Pattern Recognition: Methods that Search for Structures in Data, vol. 13, no. 8, pp. 841-847, 1991.
連結至畢業學校之論文網頁點我開啟連結
註: 此連結為研究生畢業學校所提供,不一定有電子全文可供下載,若連結有誤,請點選上方之〝勘誤回報〞功能,我們會盡快修正,謝謝!
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top
無相關論文
 
無相關期刊
 
無相關點閱論文