具視覺能力之鞦韆機器人

A Swing Robot with Vision Ability

劉士彰(Shih-Chang Liu)

指導教授 ： 王文俊

國立中央大學/資訊電機學院/電機工程學系/碩士(2007年)

摘要

本論文主要是設計一個有視覺的機器人，他會盪鞦韆，也能同時進行影像辨識。其中鞦韆運動包含了可以控制擺動到任意指定的擺角，假若有外在的干擾也能克服並達到指定的擺角；機器人不但在靜止時能夠辨識目標（綠球），在盪鞦韆時也能搜尋目標並判斷目標的距離，以能在正確時機及正確位置成功打擊目標。在機器人的機構上我們以壓克力與AI馬達組合而成身體與手腳，使用一台USB網路攝影機做為頭部，個人電腦透過RS-232協定封包來控制AI馬達的角度，並偵測其角度的回傳值以調整控制參數。在控制策略上，我們控制機器人的頭去做搜尋與辨識目標，控制機器人的身體和腳讓鞦韆擺盪到指定的擺角，控制機器人的手在適當時機來擊打目標；在影像上則是使用DirectShow 技術使影像的處理流程達到最快。因為鞦韆運動是屬於快速的運動，為了達到精準控制的要求，因而本論文使用了不少的非同步執行程序。要如何有效率的整合這些程序並降低系統的複雜性及出錯率，此外並能夠達成所有的控制目的是此研究的最大成就。

關鍵字

視覺 ； 盪鞦韆 ； 機器人

並列摘要

The goal of this thesis is to design a robot with sight which can play swing and recognize image simultaneously. In the swing motion, the robot can be controlled to swing to any desired degree even there is some disturbance existing in the swing. The robot not only can recognize the target (green ball) when it is still, but also can search the target and estimate the distance between the target in the swing such that the robot can attack the target at the right time and at the right position. The robot is made up by eleven AI motors and those AI motors are connected by acrylic boards. Further, there is a USB camera as its head and eye. We can control the rotating degree of the motor by PC through RS-232 protocol and detect the degree of the motor from feedback for adjust the controlling parameters. In control strategy, we control the head of the robot to search and recognize the target, control the robot’s body and foot to swing to the desired degree, and control the hand to attack the target accurately in the swing motion. Moreover, DirectShow technology is used to speed up the image process. In order to reach the satisfactory control performance in the fast swing motion, several asynchronous programs are used in this thesis. Integrate those programs efficiently, reduce the complexity and mistakes of the system and achieve the control objectives are the greatest accomplishments of this thesis.

並列關鍵字

vision ； swing ； robot

參考文獻

[1] K. Nagasaka, A. Konno, M. Inaba, and H. Inoue, “Acquisition of visually guided swing motion based on genetic algorithms and neural networks in two-armed bipedal robot,” IEEE International Conference on Robotics and Automation, vol.4, pp.2944-2949, April 1997.

[2] M. Inaba, K. Nagasaka, F. Kanehiro, S. Kagami, and H. Inoue, “Real-time vision-based control of swing motion by a human-form robot using the remote-brained approach,” IEEE/RSJ International Conference on Intelligent Robots and Systems, vol.1, pp.15-22, Nov. 1996.

[3] T. Yamada, K. Watanabe, K. Kiguchi, and K. Izumi, “Acquisition of exercise skill represented by fuzzy control rules for a rings gymnastic robot,” IEEE International Symposium on Computational Intelligence in Robotics and Automation, pp.368-373, 2001.

[4] T. Yamada, K. Watanabe, K. Kiguchi, and K. Izumi, “Acquiring performance skill of backward giant circle by a rings gymnastic robot,” IEEE International Conference on Robotics and Automation, vol.2, pp.1565-1570, 11-15 May 2002.

[6] S. Nakaoka, A. Nakazawa, K. Yokoi, H. Hirukawa, and K. Ikeuchi, “Generating whole body motions for a biped humanoid robot from captured human dances,” IEEE International Conference on Robotics and Automation, vol.3, pp.3905-3910, 2003.