本論文設計實現一個小型人形機器人的行走訓練平台，其可以對一個在長×寬為300(公分)×120(公分)場地上的小型人形機器人分別做定位與追蹤，並且可以對人形機器人進行升降與移動的動作。在平台系統的設計與實現上，主要有三項：(1)視覺為基礎的定位系統，(2)通訊與策略系統，以及(3)模糊控制系統。在視覺為基礎的定位系統之設計與實現上，本系統分別在平台與機器人上面放置一些定位色塊與機器人色塊，並且使用二個攝影機來擷取這些色塊的影像資訊，以及設計一個視覺系統來對平台與機器人做定位。在通訊與策略系統之設計與實現上，本系統負責傳送定位系統所得到的平台與機器人的座標，並且做策略分析來決定平台所需進行的動作。在模糊控制系統之設計與實現上，本系統在X軸方向移動與Y軸方向移動分別提出一個兩輸入一輸出的模糊控制器來控制Ｘ軸方向與Ｙ軸方向的馬達，讓平台具有X軸與Y軸移動的能力，並且對人形機器人的行走做自主的目標追蹤移動。最後，由定位誤差與小型人形機器人移動的追蹤實驗結果可知，本論文所設計實現的小型人形機器人的行走訓練平台，確實可以有效的應用在小型人形機器人的行走訓練上。

In this thesis, a walking training platform for a small-sized humanoid robot is designed and implemented. It also can locate and track the small-sized humanoid robot in the field (the size of length × width is 300 cm × 120 cm). Moreover, it can raise and lower the robot. In the design and implementation of this platform, there are three main parts: (1) vision-based localization system, (2) communication and strategy systems, and (3) fuzzy control system. In the design and implementation of the vision-based localization system, some color blocks are placed on the platform and robot. Moreover, two cameras are used to extract the image information of these color blocks and a vision system is designed to obtain the locations of the platform and the robot. In the design and implementation of the communication and strategy systems, the communication system is designed to send the coordinates of the platform and robot. The strategy system is designed to do strategic analysis and decide the motion of the platform. In the design and implementation of the fuzzy control system, a two-input one-output fuzzy control system in the X-axis direction and Y-axis direction is respectively designed to control the motor speed in the X-axis direction and Y-axis direction. Finally, some experimental results are presented to show that the implemented platform can indeed effectively be used in the walking training of a small-sized humanoid robot.

中文摘要	I
英文摘要	II
致　謝	II
目錄	III
圖目錄	V
表目錄	IX
第一章 緒論	1
1.1 研究背景	1
1.2 研究動機	3
1.3 論文架構	4
第二章 訓練平台之設計與實現	5
2.1 訓練平台開發軟硬體	5
2.2 訓練平台機構	6
2.2.1 X軸移動	7
2.2.2 Y軸移動	8
2.2.3 人形機器人升降	9
2.3 訓練平台追蹤的人形機器人介紹	12
2.4 訓練平台系統架構	13
第三章 影像處理與定位系統	16
3.1 影像定位環境與定位色塊	16
3.1.1 實驗環境	16
3.1.2 X軸定位色塊設計	18
3.1.3 Y軸定位色塊設計	20
3.1.4 RGB色彩模型	21
3.1.5 HSV色彩模型	22
3.2 X與Y軸視覺系統架構	23
3.2.1 影像擷取	24
3.2.2 影像二值化	25
3.2.3 物件標記	28
3.2.4 影像色彩分割	30
3.2.5 定位色塊的關係與辨識	31
3.3 平台與人形機器人定位系統	33
第四章 通訊與策略系統	36
4.1 平台狀態轉換	36
4.2 測試模式	36
4.3 追蹤模式	38
4.4 重新模式	39
4.5 自動模式	40
第五章 模糊控制器設計與實現	44
5.1 模糊理論介紹	44
5.2 X軸與Y軸方向移動的模糊控制器	46
第六章 實驗結果	51
6.1 定位實驗	51
6.2 追蹤誤差實驗	54
6.2.1 使用線性追蹤誤差實驗	55
6.2.2 使用模糊控制追蹤誤差實驗	57
6.3 人形機器人追蹤實驗	58
第七章 結論與未來展望	63
參考文獻	64

圖目錄
圖2.1、平台使用的硬體元件(a) The Vaio Z Series、(b) Rx-64、(c) LifeCam Cinema	6
圖2.2、平台設計圖	7
圖2.3、平台X軸移動的機構(a)設計圖、(b)實體完成圖	8
圖2.4、平台Y軸移動的機構以及Y軸的攝影機(a)使用3D繪圖軟體所繪製的機構、(b)實體完成圖	9
圖2.5、平台Y軸的吊掛機構(a)使用3D繪圖軟體所繪製的吊掛機構、(b)吊掛機構實體完成圖	10
圖2.6、人形機器人吊掛尼龍繩的機構	11
圖2.7、升降功能(a)升起人形機器人、(b)下降釋放人形機器人	11
圖2.8、訓練平台追蹤機器人	12
圖2.9、系統架構方塊圖	13
圖3.1、訓練平台座標系	17
圖3.2、人機器與NXT位於追蹤平台上的座標系	18
圖3.3、X軸色塊的配置與分別代表的座標	19
圖3.4、Y軸色塊的配置與分別代表的座標	20
圖3.5、RGB色彩模型	22
圖3.6、HSV色彩模型	23
圖3.7、影像系統處理流程圖	24
圖3.8、攝影機所擷取到的畫面	25
圖3.9、影像建模介面(a)色彩模型HSV色彩數值調整結果、(b)擷取的原始影像、(c)經過二值化處理後的結果	27
圖3.10、訓練平台X與Y軸色彩二值化結果	28
圖3.11、物件標記的遮罩(a)四鄰點遮罩(b)八鄰點遮罩	29
圖3.12、連通物件標記法的示意圖	30
圖3.13、影像分割的結果	31
圖3.14、訓練平台的影像系統介面：(a)攝影機設定、(b)錄影設定、(c)影像建模、(d)顯示影像處理結果、(e)平台參數表、(f)平台動態顯示介面、(g)影像顯示介面、(h)顯示錄影時間控制	32
圖3.15、定位的影像資訊示意圖(a)X軸定位色塊、(b)影像中心與X軸定位色塊距離、(c)Y軸定位色塊、(d)影像中心與Y軸定位色塊距離	34
圖4.1、訓練平台測試模式示意圖	37
圖4.2、訓練平台測試模式實際移動路徑圖	37
圖4.3、訓練平台的可移動範圍	39
圖4.4、人形機器人於訓練平台重新模式示意圖	40
圖4.5、狀態流程圖	43
圖5.1、模糊控制器的基本結構	44
圖5.2、X軸方向移動的雙輸入一輸出模糊控制器	46
圖5.3、Y軸方向移動的雙輸入一輸出模糊控制器	47
圖5.4、X軸模糊控制器雙輸入的歸屬函數(a)errorx的歸屬函數、(b)speedx的歸屬函數	48
圖5.5、Y軸模糊控制器雙輸入的歸屬函數，(a)errory的歸屬函數、(b)speedy的歸屬函數	48
圖5.6、X軸追蹤控制器的輸出motor_speedx的歸屬函數	49
圖5.7、Y軸追蹤控制器的輸出motor_speedy的歸屬函數	49
圖6.1、訓練平台記錄NXT原地旋轉的路徑	52
圖6.2、定位分析的路徑與實際的誤差量，平均誤差量為0.1439公分	52
圖6.3、訓練平台記錄NXT旋轉並前進以及後退的路徑	53
圖6.4、定位分析的路徑與實際的誤差量，平均誤差量為0.4029公分	53
圖6.5、訓練平台記錄NXT旋轉並前進以及後退的路徑	54
圖6.6、NXT移動的路徑(a)X軸移動的變化量、(a)Y軸移動的變化量	55
圖6.7、訓練平台使用線性的追蹤運算資料(a)X與Y軸馬達速度的變化量、(b)訓練平台與機器人的距離差	56
圖6.8、模糊控制器輸出入變化量(a)X軸輸入、(b)Y軸輸入、(c)X與Y軸輸出	57
圖6.9、使用模糊控制器追蹤NXT移動過程的誤差量	58
圖6.10、人形機器人移動的路徑	59
圖6.11、模糊控制器輸出入變化量(a)X軸輸入、(b)Y軸輸入、(c)X與Y軸輸出	60
圖6.12、人形機器人移動的路徑	61
圖6.13、訓練平台對人形機器人行走從追蹤模式到重新模式的實驗結果	62
 
表目錄
表3.1、顏色模型之色彩代碼表	26
表4.1、平台溝通封包規格(a)平台所發出的訊號格式、(b)機器人發出的訊號格式	41
表4.2、平台控制命令表(a)訓練平台的控制命令、(b)人形機器人的控制命令	42
表5.1、X軸模糊控制器的規則庫	50
表5.2、Y軸模糊控制器的規則庫	50

參考文獻
[1]	K. Hirai, "Current and future perspective of Honda humanoid robot," IEEE/RSJ International Conference on Intelligent Robots and Systems, pp. 500-508, 1997.
[2]	K. Hirai, "The development of Honda humanoid robot," IEEE InternationalConferenceon Robotics and Automation, pp. 1321-1326, 1998.
[3]	J. I. Yamaguchi, A. Takanishi, and I. Kato, "Development of a biped walking robot compensating for three-axis moment by trunk motion," IEEE/RSJ International Conference on Intelligent Robots and Systems, pp.561-566, 1993.
[4]	H. O. Lim, Y. Kaneshimat and A. Takanishi, "Online walking pattern generation for biped humanoid robot with trunk," IEEE International Conference on Robotics and Automation, pp.3111-3116, 2002.
[5]	I. W. Park, J. Y. Kim, J. Lee, and J. H. Oh, "Online free walking trajectory generation for biped humanoid robot KHR-3(HUBO)," IEEE International Conference on Robotics and Automation, pp. 1231-1236, 2006.
[6]	I. W. Park, J. Y. Kim, and J. H. Oh, "Online biped walking pattern generation for humanoid robot KHR-3(KAIST humanoid robot - 3: HUBO)," International Conference on Humanoid Robots, pp. 398-403, 2006.
[7]	Y. Ogura and H. Aikawa, "Development of a humanoid robot having 2-DOF waist and 2-DOF trunk," IEEE International Conference on Robotics and Automation, pp 134-139, 2005.
[8]	A. M. M. Omer, H. O. Lim, A. Takanishi, and H. Kondo, "Development of walking support system based on dynamic simulation," IEEE International Conference on Robotics and Biomimetics, pp. 22-25, 2009.
[9]	A. M. M. Omer, H. O. Lim, A. Takanishi and H. Kondo, "Simulation of walking support system using humanoid robot," IEEE/SICE International Symposium on System Integration, pp. 48-53, 2008.
[10]	Y. Ogura,H. Aikawa, K. Shimomura, H. Kondo, A. Morishima, H. O.  Lim, and A. Takanishi, "Development of a humanoid robot capable of leaning on a walk-assist machine," IEEE/RAS-EMBS International Conference on Biomedical Robotics and Biomechatronics, pp. 835-840, 2006.
[11]	H. Kondo, Y. Ogura, H. Aikawa, K. Shimomura, A. Morishima, H. O.  Lim, A. Takanishi, "Biped humanoid robot capable of being used as human motion simulator," The 10th International Symposium on Experimental Robotics, pp. 523-531, 2008.
[12]	J. Or, "Computer simulations of a humanoid robot capable of walking like fashion models," IEEE Transactions on Systems, Man, and Cybernetics, vol. 42, no. 2, pp.241-248, 2012.
[13]	H. J. Kang, K. Hashimoto, K. Nishikawa, E. Falotico, H. O. Lim, A. Takanishi, C. Laschi, P. Dario, A. Berthoz, "Biped walking stabilization on soft ground based on gait analysis," IEEE RAS & EMBS International Conference on Biomedical Robotics and Biomechatronics, pp. 669-674, 2012.
[14]	M. Betke and L. Gurvits, "Mobile robot localization using landmarks," IEEE Transactions on Robotics and Automation, vol. 13, no. 2, pp. 251-263, 1997.
[15]	G. Jang, S. Kim, W. Lee, and I. Kweon, "Color landmark based self-localization for indoor mobile robots," IEEE International Conference on Robotics and Automation, pp 1037-1042, 2002.
[16]	N. Ayache and O. D. Faugeras,"Building, registrating, and fusing noisy visual maps," The International Journal of Robotics Research, vol. 7, no. 6, pp 45-65, 1988.
[17]	S. F. Chen,"An on-line biped mini-robot motion learning using neural network and database management," Seventh International Conference on Natural Computation, pp 636-640, 2011.
[18]	W. T. Miller, "Real-time neural network control of a biped walking robot," IEEE Control Systems, vol. 14, no. 1, pp. 41-48, 1994.
[19]	J. G. Juang, "Fuzzy neural network approaches for robotic gait synthesis," IEEE Transactions on Systems, Man, and Cybernetics, vol. 30, no. 4, pp. 594-601, 2000.
[20]	J. G. Juang, "Trajectory synthesis based on different fuzzy modeling network pruning algorithms," IEEE International Conference on Control Applications, pp. 315-320, 2000.
[21]	J. P. Ferreira, M. Crisostomo, A. P. Coimbra, and B. Ribeiro, "Simulation control of a biped robot with support vector regression," IEEE International Symposium on Intelligent Signal Processing, pp. 1-6, 2007.
[22]	L. A. Zadeh, “Fuzzy sets, ” Information and Computation/information and Control - IANDC , vol. 8, no. 3, pp. 338-353, 1965
[23]	R. C. Luo and T. M. Chen, “Target tracking by grey prediction theory and look-ahead fuzzy logic control, ” IEEE International Conference on Robotics and Automation, pp. 1176-1181, 1999.
[24]	T. H. S. Li, S. J. Chang, and W. Tong, “Fuzzy target tracking control of autonomous mobile robots by using infrared sensors, ” IEEE Transactions on Fuzzy Systems, vol. 12, no. 4, pp. 491-501, 2004.
[25]	W. Tong and T. H. S. Li, "Realization of two-dimensional target tracking problem via autonomous mobile robots using fuzzy sliding mode control," Annual Conference of the IEEE Industrial Electronics Society, pp. 1158-1163, 1998.
[26]	R. J. Wai and Y. W. Lin, "Adaptive moving-target tracking control of vision-based mobile robot via dynamic petri recurrent-fuzzy-neural-network," IEEE Transactions on Fuzzy Systems, vol. 21, no. 4, pp. 688-701, 2012.
[27]	URL: www.robotis.com/