臺灣博碩士論文加值系統

一般的機械臂通常是在固定的環境下執行事先規劃好的工作，為了使機械臂工作更具彈性，最直接的方式就是結合視覺系統，利用影像回授來引導機械臂執行任務。而採用主動式視覺系統(active vision system)，將攝影機架設在伺服機構上，則視覺感知將更具實用性。因此本文目的即在提出一個整合機械臂與主動式視覺系統之控制架構，利用主動式視覺系統所得之影像資訊，來引導機械臂追蹤目標物。
本文以傳算器平行處理的特性來整合機械臂與主動式視覺系統。利用SOIM類神經網路來學習影像資訊與雙眼機械頭控制命令間的關係，配合卡爾曼濾波器對於空間目標物進行視覺追蹤；並透過柯漢寧類神經網路來學習空間表示式變化與機械臂控制命令間的關係，進而驅動機械臂來追蹤此目標物。
最後，將以雙眼機械頭與U型機械臂所建構而成的機械臂追蹤系統，驗證本文所提控制架構之可行性。

With the feedback of visual signals, a robot manipulator will function more independently and more versatilely. For examples, it can work in an imprecisely calibrated environment and also adapt itself to the changing-environment. Active vision is a good way to imitate human visual behavior through an image servomechanism and stereo cameras. In this thesis, a novel control structure is proposed for visual servoing of a robot manipulator with an active vision system. With the utilization of this control structure, a robot manipulator can be guided to track a moving target by an active vision system.
The goal of this thesis is to develop a framework for visual servoing of robot manipulators. The design framework is based upon the proposed control structure and involves the application of SOIM neural network, Kalman filter, and Kohonen neural network. SOIM neural network is utilized to learn a calibration-free spatial representation of 3D targets. Kohonen neural network is utilized to learn the mapping from incremental changes in the spatial representation to incremental changes in the joint configurations of a robot manipulator. In order to predict the trajectory of a moving target, Kalman filter is used.
Finally, the theoretical results are experimentally verified on an industrial robot and an active vision system, which is set up at the Robotic Systems and Control Laboratory of National Cheng Kung University. Specific experimental demonstrations include active visual tracking of the active vision system and visual servoing of the robot manipulator for a stationary and moving target, respectively.

摘要
英文摘要
誌謝
目錄
圖目錄
表目錄
第一章 緒論
1.1前言
1.2文獻回顧
1.3研究動機與目的
1.4本文架構
第二章 雙眼機械頭之動力學推導
2.1座標系統
2.2運動學推導
2.3雙眼機械頭連桿的動態方程式
2.4雙眼機械頭連桿-致動器的動態方程式
第三章 雙眼機械頭之控制器設計
3.1滑動模式控制
3.2模糊滑動模式控制
3.3雙眼機械頭之模糊滑動模式控制器設計
第四章 機械臂追蹤系統控制器之設計
4.1整體系統之控制架構
4.2視覺追蹤控制器之設計
4.2.1存在唯一性
4.2.2 SOIM類神經網路
4.3柯漢寧類神經網路
4.4卡爾曼濾波器
第五章 機械臂追蹤系統之硬體架構
5.1雙眼機械頭子系統
5.2 U型機械臂子系統
5.3 傳算器網路架構
5.4 視覺子系統
5.4.1CCD攝影機
5.4.2C4F網路
5.4.3影像處理
5.5完整實驗架構
第六章 實驗
6.1雙眼機械頭之定位控制
6.2SOIM類神經網路學習結果
6.3主動式視覺追蹤實驗
6.3.1靜態目標物
6.3.2動態目標物
6.4視覺導引機械臂追蹤實驗
6.4.1靜態目標物
6.4.2動態目標物
第七章 結論
7.1結果與討論
7.2未來展望
參考文獻
附錄A
附錄B
自述
著作權聲明

[1]. John Aloimonos, Isaac Weiss, and Amit Bandyopadhyay, "Active Vision," International Journal of Computer Vision, vol. 1, pp. 333-356, 1988.
[2]. H. Asada and Jean-Jacques E. Slotine, "Robot Analysis and Control," Wiley. Interscience, 1986.
[3]. Ruzena Bajcsy, "Active Perception," Proceeding of the IEEE, vol. 78, pp. 996-1005, 1988.
[4]. Robert G. Brown, Patrick Y.C. Hwang, "Introduction to random signals and applied Kalman filtering," Wiley, 1992.
[5]. G. A. Carpenter, S. Grossberg, and D. B. Rosen, "Fuzzy ART : Fast Stable Learning and Categorization of Analog Patterns by an Adaptive Resonance System," Neural Networks, vol. 4, pp. 759-771, 1991.
[6]. P. Gaudiano and S. Grossberg, "Vector associative maps: Unsupervised real-time error-based learning and control of movement trajectories," Neural Networks, vol. 4, pp. 147-183, 1991.
[7]. W. Jesse Hong, "Comparison of General Methods for Tracking a Moving Object," Technical Report, Massachusetts Institute of Technology, May 1997.
[8]. W. Jesse Hong and Jean-Jacques E. Slotine, "Experiments in Hand-Eye Coordination Using Active Vision," Experimental Robotics IV. 4th International Symposium, pp. 130-139, 1997.
[9]. Seth Hutchinson, Gregory D. Hager, and Peter I. Corke, "A Tutorial on Visual Servo Control," IEEE Transactions on Robotics and Automation, vol. 12, pp. 651-670, 1996.
[10]. Paul R. Kalata, "The Tracking Index: A Generalized Parameter for and Target Trackers", IEEE Transactions on Aerospace and Electronic Systems," vol. 2, pp. 174-182, 1984.
[11]. Brian C. Madden and Ulf M. Cahn von Seelen, "PennEyes - A Binocular Active Vision System," Technical Report MS-CIS-95-37, University of Pennsylvania, December 1995.
[12]. A. C. Sanderson and L. E. Weiss, "Image-based Visual Servo Control Using Relational Graph Error Signals," Proceeding of IEEE, pp. 1074-1077, 1980.
[13]. Jean-Jacques E. Slotine, "Sliding controller design for non-linear systems," International Journal of Control, vol. 40, pp. 421-434, 1984.
[14]. Jean-Jacques E. Slotine and Weiping Li, "Applied Nonlinear Control," Prentice Hall, 1991.
[15]. Narayan Srinivasa and Rajeev Sharma, "Efficient Learning of VAM-Based Representation of 3D Targets and its Active Vision Applications," Neural Networks, vol. 11, pp. 153-171, 1998.
[16]. Narayan Srinivasa and Rajeev Sharma, "SOIM: A Self-Organizing Invertible Map with Applications in Active Vision," IEEE Transactions on Neural Netowrks, vol. 8, pp. 758-773, 1997.
[17]. Narayan Srinivasa and Rajeev Sharma, "A framework for active vision-based robot control using neural networks," Robotica, vol. 16, pp. 309-327, 1998.
[18]. Greg Welch and Gary Bishop, "An Introduction to the Kalman Filter," Technical Report 95-041, University of North Carolina at Chapel Hill, October 1998.
[19]. 葉怡成，"類神經網路 - 模式應用與實作"，儒林圖書有限公司，民國八十二年。
[20]. 黃健華，"模糊滑動模式控制應用於機械臂系統之研究"，成大機械碩士論文，民國八十三年。
[21]. 戴文彬，"傳算器架構之雙機械臂智慧型協調控制"，成大機械碩士論文，民國八十四年。
[22]. 李育忠，"傳算器架構之長控機器人系統智慧型控制"，成大機械碩士論文，民國八十四年。
[23]. 李國駿，"雙眼機械頭之設計及其在視覺伺服之應用"，成大機械碩士論文，民國八十七年。