機械手臂結合影像系統之控制

Mechanical Arm Control Combined with Image System

葉傅文(Fu-Wen Ye)

指導教授 ： 陳美勇

國立臺灣師範大學/科技學院/機電科技研究所/碩士(2012年)

摘要

本論文的研究內容為使用機械手臂結合影像辨識系統，取得工作空間中目標物件之座標，以進行物件的抓取或移動。由於機械手臂在現實生活當中的應用存在許多變數，不同的任務下針對物件姿態所能容許的移動方式可能有所限制，例如移動盛水的杯子要避免傾倒的姿勢。一般過去的研究僅強調物件定位的精確度，而並未考慮機械手臂的姿態，有鑒於此，本控制系統會在執行物件的抓取時，依據任務之目的切換不同的控制策略，以符合正確的任務目的與物件擺放姿態。
若要將機械手臂整合影像系統並成功應用於實作，則必須依照工作空間內的變化做出即時的運算，本研究除了利用影像處理進行物件的輪廓與顏色判別外，還配合夾爪上的雷射光模組所投影的光點作為回饋進行定位。在本研究當中所使用的機械手臂具有六軸關節存在運動學冗餘度的問題，因此本研究之系統必須事先進行D-H座標系統的順向與逆向運動學分析，推算出三維空間卡式座標系統與機械手臂各關節馬達轉動角度之間的關係，如此一來才能實現快速、靈活與準確的控制。本研究最後成功建立一套通用的多軸機械手臂控制方法，能夠應用到各種類似配置的機械手臂上，透過影像處理分析攝影機接收到的資訊，以應付各種不同的環境下更加複雜的應用與操作。

關鍵字

機械手臂 ； 影像辨識 ； D-H 座標系統 ； 逆向運動學

並列摘要

In this paper, we propose a general approach to control mechanical arm which combine an image identification system. Obtain the object’s coordinate information in workspace, in order to move an object to a desire position. There are many situations when using mechanical arm in reality applications, the object's posture may be restricted in different tasks. Many past studies only emphasized the accuracy of object location, but did not consider the posture of the robot arm. Therefore, our control system will switch to different control strategy according to the purpose of tasks.
Mechanical arm has 6-DOF and can perform highly flexible action, analysis forward and backward kinematics equations from D-H coordinate system. After computed object coordinate, IK solution methods are applied to mechanical arm gripper position control. Finally, the mechanical arm can distinguish between difference figures and colors. Our study hopes to establish a general multi-axis mechanical arm control method which can be applied to mechanical arm with similar configuration. Analysis camera information received through the image processing to meet the more complex applications and operating in different environments.

並列關鍵字

Mechanical arm ； Vision classification ； IK solution ； D-H coordinate system

參考文獻

[1] L. W. Sun, F. Van Meer, Y. Bailly, C. K. Yeung, “Design and Development of a Da Vinci Surgical System Simulator,” IEEE, International Conference on Mechatronics and Automation, pp. 1050-1055, 2007.

[2] J. Dastur, A. Khawaja, “Robotic Arm Actuation with 7 DOF Using Haar Classifier Gesture Recognition,” IEEE, International Conference on Computer Engineering and Applications, vol. 1, pp. 27-29, 2010.

[3] D. Tolani, A. Goswami and N. Badler, “Real-time inverse kinematics techniques for anthropomorphic limbs,” Graph. Models, vol. 62, pp. 353-388, 2000.

[4] D. E. Whitney, “The mathematics of coordinated control prosthetic arm and manipulators,” Trans. ASME J. Dyn. Syst., Meas. Control, vol. 94, pp. 303-309, 1972.

[5] De Lope, R. Gonzalez-Careaga and T. Zarraonandia, “Inverse kinematics of humanoid robots using artificial neural networks,” in Proc. Int. Workshop Comput. Aided Syst. Theory, pp. 216-218, 2003.