臺灣博碩士論文加值系統

控制大型雙足人形機器人是一項相當有挑戰且困難的任務，因為機構剛性、背隙以及地面的環境等未知因素都可能造成機器人跌倒。故此，本論文提出一以多重感測器回授為基礎之機器人步態穩定控制系統，來增加機器人在步態中不確定因素的適應能力。本系統由兩組壓力感測單元以及一個九軸慣性感測單元（Inertial measurement unit; IMU）組成，其中壓力感測單元用來測量機器人行走時足底壓力中心（Center of pressure; COP）的軌跡，並以此計算出步態中的壓力中心軌跡所構成之面積及該面積的中心點（Center of Area; COA）。而慣性感測單元中的陀螺儀則用來量測機器人每個步態週期中的髖部平面Roll軸向的角速度並算出「基於半步態週期之陀螺儀穩定度指標」（Half Cycle-base Gyro Stable Index; HCBGSI）。
為使機器人能夠適應不同的地面環境，如斜坡、不平坦地面等，機器人開始步行之前會先以壓力感測單元及IMU的資訊調整機器人各關節的位置，使其達到初始條件，其定義如下：機器人之左右腳踝均服貼於地面、COP的位置在機器人座標系原點（機器人兩足的中心），以及機器人髖部平面為水平狀態。當機器人開始步行之後，修正分為週期性修正和即時性修正兩部分。在週期性修正的部分，COA以及HCBGSI分別會被用來計算出髖部的擺盪中心偏移及擺盪振幅的修正量，並在下一個跨步時做出修正。而即時性修正的部分則是針對陀螺儀所量測之髖部平面在Roll軸向的角速度，經另一比例微分控制器控制器做出即時修正，以確保機器人在步行時遇到週期性修正不足以修正機器人時仍能可維持其穩定性。
最後，本研究以HuroEvoutionAD 大型雙足人形機器人作為驗證平台，測試其往前、後、左、右以及全方位步行的能力。經測試後，可使該機器人穩定行走於14%斜坡及以每步30公分的快速穩定行走。相較於加入此系統之前，其坡度適應能力提升了兩倍，而快速行走速度則提升40%。

Controlling an adult-size humanoid robot is a challenging and difficult task due to mechanical rigidity, backlash, ground environment and other unknown factors which may cause the robot to fall. Therefore, this thesis proposed a gait stability control system for humanoid robots based on multi-sensory feedback, in order to deal with the uncertainty. The system is consist of two pressure sensing units and a 9-axis inertial measurement unit (IMU). The pressure sensing unit is used to measure center of pressure (COP) of the foot during the robot is walking. The area as well as the center of the area (COA) formed by the COP trajectory in a stride is then calculate. In the other hand, the gyroscope in the IMU is used to measure the angular velocity along the roll axial, which is then calculated as half cycle-base gyro stable index (HCBGSI).
In order to adapt to different environment, such as slopes, uneven ground, etc., the position of each joints will be adjust based on the information from pressure sensing units and the IMU to let the robot reach the initial conditions before the robot performing a stride. The initial condition is defined as follows: the left and the right ankles of the robot are parallel to the ground, the COP locate in origin of the robot coordinate and the hip plane of the robot is horizontal. After the robot began to walk, the adjustment is divided into two parts: a cycle-base adjustment and a real-time adjustment. In the cycle-base adjustment, COA and CBGSI will be used to calculate the adjustment of the bias and amplitude of hip swinging. In addition, all corrections in the cycle-base adjustment would be performed in next stride. As for the real-time adjustment, the correction is made based on the angular velocity along the roll axial measured from the gyroscope. Since it is a real-time adjustment, the correction is then output to the hip plane immediately to increase the stability.
Finally, the adult size platform HuroEvoutionAD is used in this study. After the experiment, the robot was able to perform fast walking with 25cm per footstep and stable walking on a 15% slope. Compared to the original system, the ability of robot walking on slope was enhanced twice times, and the walking speed become 40% faster after implementing the proposed system.

指導教授推薦書 I
委員審定書 II
授權書 III
誌謝 IV
中文摘要 VI
ABSTRACT VII
目錄 VIII
圖目錄 XI
表目錄 XIV
參數對照表 XV
第1章 緒論 1
1.1 研究背景與動機 1
1.2 論文架構 3
第2章 文獻回顧 4
2.1 應用於雙足機器人之感測器相關文獻探討 4
2.2 雙足機器人軌跡生成相關文獻探討 6
2.3 應用於雙足機器人之控制器相關文獻探討 7
2.4 文獻總結 8
第3章 實驗平台介紹 9
3.1 機器人機械結構 9
3.2 自由度配置 11
3.3 機器人控制架構 12
3.3.1 機器人致動器 13
3.3.2 機器人動作控制器 14
3.3.3 髖部軌跡生成 15
3.3.4 雙足末端點軌跡生成 17
3.3.5 機器人逆向運動學推導 18
第4章 研究方法 22
4.1 系統架構 22
4.2 機器人感測器設計 23
4.2.1 足底壓力感測單元 23
4.2.2 慣性感測單元 25
4.2.3 感測器精度測試 25
4.3 人類步行相位分析 26
4.4 壓力中心點計算 29
4.5 機器人穩定性指標及控制 30
4.5.1 基於步態週期之壓力感測單元穩定度指標 32
4.5.2 基於步態週期之陀螺儀穩定度指標 34
4.5.3 基於即時性之髖部平面修正 36
第5章 實驗結果 38
5.1 足底壓力感測器校正測試 38
5.2 機器人初始姿態態校正 41
5.3 機器人步態相位閥值設定 46
5.4 基於步態週期之壓力感測單元穩定控制試驗 50
5.5 基於步態週期之陀螺儀穩定控制實驗 52
5.6 基於即時性之髖部平面修正實驗 55
5.6.1 髖部平面重力補償 56
5.6.2 即時性之髖部平面修正 57
第6章 結論與未來研究方向 61
6.1 結論 61
6.2 未來研究方向 65
參考文獻 66

[1]M. Vukobratovic, “ZERO-MOMENT POINT — THIRTY FIVE YEARS OF ITS LIFE, ” International Journal of Humanoid Robotics, vol. 1, no. 1, pp. 157 – 173, 2004.
[2]S. Kajita and K. Tani, “Experimental Study of Biped Dynamic Walking in the Linear Inverted Pendulum Mode,” IEEE International Conference on Robotics and Automation, pp. 2885–2891, 1965.
[3]J. H. Park and E. S. Kim, "Foot and Body Control of Biped Robots to Walk on Irregularly Protruded Uneven Surfaces", IEEE Trans. Syst., Man, Cybern. B, vol. 39, no. 1, pp. 289-297, 2009.
[4]A. Kalamdani, C. Messom and M. Siegel, "Robots with Sensitive Feet", IEEE Instrum. Meas. Mag., vol. 10, no. 5, pp. 46-53, 2007.
[5]E. Hwang, J. Seo and Y. Kim, "A Polymer-Based Flexible Tactile Sensor for Both Normal and Shear Load Detections and Its Application for Robotics", Journal of Microelectromechanical Systems, vol. 16, no. 3, pp. 556-563, 2007.
[6]J. S. Hu, K. C. Sun and C. Y. Cheng, "A Kinematic Human-Walking Model for the Normal-Gait-Speed Estimation Using Tri-Axial Acceleration Signals at Waist Location", IEEE Transactions on Biomedical Engineering, vol. 60, no. 8, pp. 2271-2279, 2013.
[7]C. L. Shih, "Ascending and descending stairs for a biped robot", IEEE Transactions on Systems, Man, and Cybernetics - Part A: Systems and Humans, vol. 29, no. 3, pp. 255-268, 1999.
[8]Q. Huang and K. Yokoi, “Planning walking patterns for a biped robot,” IEEE Transaction on Robotics and Automation, vol. 17, pp. 280-289, 2001.
[9]T. Hemker, M. Stelzer, O.V. Stryk and H. Sakamoto, “Efficient walking speed optimization of a humanoid robot,” International Journal Robotic Research, pp. 303-314, 2009.
[10]F. Asano, M. Yamakita, N. Kamamichi and Z. Luo, "A Novel Gait Generation for Biped Walking Robots Based on Mechanical Energy Constraint", IEEE Trans. Robot. Automat., vol. 20, no. 3, pp. 565-573, 2004.
[11]Q. Huang and Y. Nakamura, “Sensory reflex control for humanoid walking,” IEEE Transactions on Robotics, vol. 21, no. 5, pp. 977-984, 2005.
[12]D. Kim, G. Park and S. Seo, "Zero-moment point trajectory modelling of a biped walking robot using an adaptive neuro-fuzzy system", IEE Proceedings - Control Theory and Applications, vol. 152, no. 4, pp. 411-426, 2005.
[13]J. Ferreira, M. Crisostomo and A. Coimbra, "SVR Versus Neural-Fuzzy Network Controllers for the Sagittal Balance of a Biped Robot", IEEE Trans. Neural Netw., vol. 20, no. 12, pp. 1885-1897, 2009.
[14]C. Mummolo, "Quantifying dynamic characteristics of human walking for comprehensive gait cycle. - PubMed - NCBI", Ncbi.nlm.nih.gov, 2016. [Online]. Available: http://www.ncbi.nlm.nih.gov/pubmed/23775488. [Accessed: 01- Jun- 2016].
[15]洪錦貴，「以步態週期為基礎之雙足機器人步行控制」，國立台灣科技大學碩士論文，民國一百年。
[16]鍾睿洲，「具重心高度調適之雙足線性倒單擺步行控制」，國立台灣科技大學碩士論文，民國一百零一年。