三爪機械手之控制驅動與感測系統之開發與應用展示

Development of control and sensing system for three-finger robot hand and demonstration

陳俊廷(Chun-Ting Chen)

指導教授 ： 楊燿州

國立臺灣大學/工學院/機械工程學研究所/碩士(2010年)

https://doi.org/10.6342/NTU.2010.01331

摘要

本文主要的目的在設計一三爪機械手的驅動控制與感測系統，使其能夠驅動控制三爪機械手各關節移動的位置及輸出的力量，達到抓握物體的目的，並且利用感測系統所收集的力量資訊，得到抓握時的實際受力情況，以及受外力影響時，能做即時的判斷，使整個抓握動作更加穩定。
在這篇論文中，我們首先針對位置及力量的馬達控制器去做設計。本研究使用微處理器組成馬達的控制器，在馬達位置控制中，我們考慮了適性的PD控制器（adaptive PD controller），而力量的控制則是利用單純的PID控制器，此外並參考了基本的力量/位置混和控制的架構將兩控制器整合在一起。接下來感測系統的的設計，是由市面上所受的壓力感測器及其它材料所組成，使其能夠有多軸向的力量感測機制，並且將其裝置在三爪機械手的指尖部分，這也就是本研究的三爪機械手所配置的感測系統。
為了使具驅動控制及感測系統的三爪機械手能夠展示其應用，本論文也在硬體及軟體方面做設計，在硬體方面，使用XY平台做虛擬的手臂，並且設計符合系統的載台能夠連結三爪機械手的整個驅動控制系統並且固定於XY平台。軟體方面，則主要利用Labview為主要架構，在Labview的程式內，除了能夠控制三爪機械手的位置及力量輸出外，還有接收並統整感測系統資訊的功用。除此之外，還可以利用Labview連接MATLAB做比較複雜的運作，如抓握物體的體論，如force-closure。

關鍵字

三爪機械手 ； 馬達控制系統 ； 觸覺感測器 ； 抓握

並列摘要

In this work, we design and implement a control and sensing system for a three-finger robot hand. The system effectively controls the joint angles and forces of the hand. Then, the grasping capability of the hand is demonstrated using the system. The MSP430f1611 is employed as the microcontroller for the position and force feedback control circuit. For the position control, the adaptive PD controller is employed, while the traditional PID controller is used for the force control. In addition, using the force/position hybrid control method, the aforementioned two controllers are integrated. We adapt the commercially-available pressure sensor elements to build a multi-axis sensing system.
The required platform for the demonstration of control and sensing system are also developed. We use an xy table to emulate the arm of a robot hand. The developed control/driving/sensing circuit boards are attached on the xy table. Labview is employed as the software platform for the high-level control commands. Furthermore, the codes developed in Labview is connected with Matlab for computing with other control algorithms, such as the force closure. Moreover, with the measured force data collected by sensing system, it is demonstrated that the hand can be real-time controlled to steadily grasping a object.

並列關鍵字

three-finger robot hand ； motor control system ； tactile ； grasp

參考文獻

[1] Ascari, L., et al., Bio-inspired grasp control in a robotic hand with massive sensorial input. Biological Cybernetics, 2009. 100(2): p. 109-128.

[2] Bernardin, K., et al., A sensor fusion approach for recognizing continuous human grasping sequences using hidden Markov models. Ieee Transactions on Robotics, 2005. 21(1): p. 47-57.

[3] Bianco, R. and S. Nolfi, Evolving the neural controller for a robotic arm able to grasp objects on the basis of tactile sensors. Ai(Asterisk)Ia 2003: Advances in Artificial Intelligence, Proceedings, 2003. 2829: p. 375-384.

[4] Bianco, R. and S. Nolfi, Evolving the neural controller for a robotic arm able to grasp objects on the basis of tactile sensors. Adaptive Behavior, 2004. 12(1): p. 37-45.

[5] Dollar, A.M., et al., Contact sensing and grasping performance of compliant hands. Autonomous Robots, 2010. 28(1): p. 65-75.