臺灣博碩士論文加值系統

本論文之主旨為提供一種以人工智慧演算法應用於焊接治夾具導線嵌入製程之動作產生系統，此系統將機器視覺整合於導螺桿滑台與治夾具系統中，可望取代傳統USB線生產製程¬與焊接工序中的人力的使用並能提高生產效率。目前市面上不同硬體間資料傳輸的方式大多使用USB線，USB線也廣泛應用於智慧型手機與3C家電的充電用途，其需求量是與日俱增。為解決供不應求的問題，企業通過與學校的合作研發出一套半自動化的生產線，以機器取代了大部分的人力，但因USB線機械性質的問題，其中焊接治夾具導線嵌入之製程工序製程在成功率與穩定度方面皆不甚理想，本論文即在研究於此部分之自動化並藉由實驗來探討其中之問題。
典型的USB線焊接工序製程分為三個部分: 首先須剝取纜線外皮而裸露出內裡紅、綠、白、金屬四條不同顏色的導線，接著將這四條導線準確嵌入焊接治具上與導線顏色相對應的固定線槽，最後再把USB頭與導線焊接在一起。其中第二部分將導線嵌入線槽部分會因導線機械性質而無法做到完全自動化。導線具有撓性，其易受外力干擾而往任意方向彎曲，再者，纜線係以纏繞方式製作，使得四條不同顏色的導線在纜線內為非固定式分佈，因而無法確定每次剝取纜線後內裡四條導線的準確位置，由於這兩項因素導致此段工序於完全自動化時產生的困難。
本研究以機器視覺經由影像處理來取得導線於工作環境之之各項物理參數，再使用以LabVIEW建立A*演算法之模擬環境與路徑規則進行模擬評估，最後使用機台對其結果進行實機驗證，觀察在相同環境下以不同規則下所模擬的最佳化路徑之間的差異，並於本文最後對驗證結果做總結並建議最適合的路徑演算規則。

Discussion of an action generation system by using the artificial intelligence algorithm to the welding fixture wire embedding process is the purpose of this thesis。The developed system integrates the machine vision into the sliding table and fixture system, which is expected to eliminate the labor cost and increase production efficiency in the traditional human USB wires embedding operation before the welding process. Nowadays, the data transmission methods between general hardware devices uses the USB cable, and the USB cable is used for charging smart phones and 3C home appliances also. The demand of USB cable is increasing in each every day. In order to solve the situation of short supply, company has developed a semi-automatic production line through project with the University to replace the manpower in this process. However, due to the mechanical properties of the USB cable, the process of soldering fixture wire insertion process shows difficulties to be fully automated.
The typical USB wire soldering process is divided into three steps: First, strip the cable sheath and expose four different colors wires (Red, Green, White, Bare metal) inside the cable. Then, accurately embedded the four wires into the welding fixture slots which are corresponding with the wires color, and finally solder the USB head to the wires. Due to the mechanical properties of the wire, the second step is difficult to be fully automated. Since wires are flexible and are easily bent by any external force into all possible directions and the different color wires are randomly distributed in the cable so it is hard to know the exact position of the four wires inside each stripped cable. These two factors make this process difficult to be fully automated.
In this study, the analysis results of vision information are used with the LabVIEW to establish simulation environment and path rule of A* algorithm. The simulation results are verified by implement on the Lab-built wire embedding system to realize the effectiveness under different rules for the action generating algorithm. At the end of the thesis, the verification results are summarized, and the most suitable path rules are suggested.

摘要 ................................................i
Abstract ...............................................iii
誌謝 ................................................v
目錄 ...............................................vi
表目錄 ..............................................viii
圖目錄 ...............................................ix
符號說明 ...............................................xii
第一章 緒論............................................1
1.1 研究背景與動機....................................1
1.2 相關研究之回顧....................................2
1.3 研究目的與方法....................................2
1.4 論文架構.........................................3
第二章 導線嵌入製程機台設計..............................4
2.1 機台設計概念.....................................4
2.2 導線嵌入製程機台設計..............................6
2.3 三軸導螺桿滑台控制系統架構.........................9
2.3.1 導螺桿滑台系統架構................................9
2.3.2 三軸導螺桿滑台解析度..............................11
2.4 焊接治具設計與製作................................13
2.4.1 焊接治具設計.....................................13
2.4.2 焊接治具製作.....................................14
第三章 夾爪控制系統.....................................16
3.1 夾爪控制系統.....................................16
3.2 電動夾爪.........................................17
3.2.1 步進馬達.........................................17
3.2.2 編碼器.......................................... 18
3.3 工作夾爪設計.....................................18
3.4 夾爪控制系統分析................................. 19
3.5 夾爪控制模式.....................................21
第四章 雙視覺影像系統...................................23
4.1 雙視覺影像系統硬體規格............................23
4.2 雙視覺影像系統校正................................24
4.3 雙視覺影像共同參考原點.............................25
4.4 焊接治具影像預處理................................26
第五章 FWEWF演算法設計...................................28
5.1 路徑演算規則.....................................28
5.2 演算模型設計.....................................29
第六章 FWEWF演算法路徑演算模擬............................32
6.1 導線狀態模型設計..................................32
6.2 導線狀態模型模擬..................................34
6.3 避障型路徑演算規則................................40
第七章 實驗結果分析與探討................................43
7.1 實驗設計.........................................43
7.2 導線狀態路徑演算實驗..............................45
7.3 避障型路徑演算規則實驗............................48
第八章 結論............................................52
參考文獻....................................................53
Extneded Abstract..........................................54

[1]吳柏勳, 吳岳恩, 林稟豐, 楊仕安, 五月 31, 2017, “四線自動排線原型機開發專題研
究”, 國立虎尾科技大學自動化工程系.
[2]深圳市中海通機器人有限公司, TYPE-C生產機器人,
http://www.szmp.com/productlist.aspx?typeid=2.
[3]Xiang Li, Xing Su, Yuan Gao, and Yun-Hui Liu, May 21-25, 2018, “Vision-
Based Robotic Grasping and Manipulation of USB Wires”, IEEE
International Conference on Robotics and Automation(ICRA), Brisbane,
Australia.
[4]Stefano Ghidoni, Matteo Finotto, Emanuele Menegatti, “Automatic Color
Inspection for Colored Wires in Electric Cables”, IEEE Transactions on
Automation Science and Engineering,p.596-607,October,2014
[5]Xiang Liu, Daoxiong Gong, April 15-17, 2011, “A Comparative Study of A-
star Algorithm for Search and Rescue in Perfect Maze”, IEEE
International Conference on Electric Information and Control
Engineering(ICEICE), Wuhan, China.
[6]Peter E. Hart, Nils J. Nilsson, Bertram Raphael, July, 1968, “A Formal
Basis for the Heuristic Determination of Minimum Cost Paths”, IEEE
Transactions on Systems Science and Cybernetics, p.100-107.
[7]Yi Zhang, Junhao Wen, Jun Zeng, Qingyu Xiong, July 10-14, 2016, “Design
Method and Information Representation for Computer Aided Welding
Fixture Design”, 5th IIAI International Congress on Advanced Applied
Informatics(IIAI-AAI), Kumamoto, Japan.
[8]H. Nakagaki, K. Kitagaki, T. Ogasawara, H. Tsukune, April 25, 1997,
“Study of deformation and insertion tasks of a flexible wire”,
Proceedings of International Conference on Robotic and Automation,
Albuquerque, NM, USA.
[9]Wei-chen Lee, Kai-siang Cao, May 6-9, 2019, “Application of Machine
Vision to Inspect Wiring Harness”, IEEE International Conference on
Industrial Cyber Physical Systems (ICPS), Taipei, Taiwan.