發展大型雙足機器人具備相當地挑戰性，除了機構設計複雜，在步態控制上也較難達到穩定且快速行走。線性倒單擺是目前較普遍之步態規劃方法，但其質心運動於同一高度平面之假設，與真實人類之步行不同。因此本論文提出一類人倒單擺步態規劃（Human-Like Inverted Pendulum Model；HLIPM），其目的在於學習人類步行特徵，透過修正線性倒單擺模型，來達成仿人步態規劃。
本論文使用動作擷取系統，記錄二十二位實驗者的行走步態，其包含原地踏步，與不同速度下之前進、後退步態，並以此資料透過平滑支撐向量迴歸（Smooth Support Vector Regression；SSVR），建立虛擬質心點高度變動量、髖平面角度變動量與上半身軀幹前後傾角之預測模型，使系統可在已知機械結構資訊與步行動作特性下，可自動調適此三項變數。
為了驗證此研究，本論文使用高146公分、重量15公斤之大型雙足人形機器人作為實驗平台，並針對原地踏步、前進、後退步態，進行穩定度與速度測試。穩定度之研究結果，在本論文所定義之足部間隙穩定度指標、側向晃動穩定度指標與髖平面角度穩定度指標上，相較於傳統線性倒單擺，穩定度至少提升16%以上。另外機器人最大步行速度可達28公分/秒（1公里/小時），相較於傳統線性倒單擺，其速度提升16.7%。本論文之演算法與機器人平台實際驗證於國際性競賽2013 Robocup大型人形機器人足球組，並獲得技術挑戰賽第一名與帶球射門第二名。

Development of adult-size humanoid robots is a very challenging research topic. In addition to complicated mechanical design, the robot developer has to deal with locomotion control for improving the walking speed and stability. Linear inverted pendulum model (LIPM) is a popular locomotion approach for controlling biped humanoid robots, and the LIPM approach is used when the robot’s CoM (center of mass) is moving on the plane with a specific height. Such a phenomenon is different to the walking characteristics of human beings. Therefore, this study proposes a human-like inverted pendulum model (HLIPM) for controlling an adult-size humanoid robot. The HLIPM is developed by investigating and emulating the walking characteristics of human beings, and then the parametric locomotion characteristics is further used to dynamically adjust the parameters used in ordinary LIPM approaches.
To collect the motion data of human beings, a motion capture system was used for experiments. Twenty-two subjects were recruited for evaluating the different locomotion, such as mark time motion and forward and backward walking with different speed. The motion data is evaluated in terms of smooth support vector regression (SSVR) approach to discover the parametric locomotion characteristics. As a consequence, the SSVR approach is capable of generating parametric locomotion characteristics for virtual CoM height, hip’s plane angle, and torso tilt angle according to different mechanical structures and locomotion commands.
To evaluate the performance of the proposed HLIPM, an adult-size humanoid robot with 146 cm tall and 15 kg in weight was used in this thesis. The experiments collected the robot's performance data from a speed-controlled powered treadmill for the locomotion patterns of mark time motion and forward and backward walking with different speed. The stability indices were evaluated with foot clearance stability index (FCSI), lateral swing stability index (LWSI) and hip plane angle stability index (HPASI). The result showed that the stabilities and speeds of walking are improved when compared to ordinary LIPM. Comparing to the LIPM results, the HLIPM improved these three stability indices for at least 16%. Moreover, the maximum stable forward walking speed reached 28 cm/sec (i.e., 1 km/hr), and it improved 11.7% from the LIPM approach. This research have been applied in the 2013 Robocup humanoid league adult size soccer competition, and won the champion of technical challenge, the second place of dribble and kick competition.

[1] M. Vukobratovic, “Zero-Moment-Point Thirty Five Years of its Life,” IEEE 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.I Yamaguchi, A. Takanishi, and I. Kato, “Development of a Biped Walking Robot Compensating for Three-Axis Moment by Trunk Motion,” IEEE International Conference on Intelligent Robotics and Systems, pp. 561 – 566, 1993
[4] E. Ohashi, T. Sato, K. Ohashi, “A Walking Stabilization Method Based on Environmental Modes on Each Foot for Biped Robot,” IEEE Transactions on Industrial Electronic, vol. 56, no. 10, pp. 3964 – 3974, 2009.
[5] B.J Lee, D. Stonier, Y.D Kim, J.K. Yoo, and J.H. Kim, “Modifiable Walking Pattern of a Humanoid Robot by Using Allowable ZMP Variation,” IEEE Transactions on Robotics, vol. 24, no. 4, pp. 917 – 925, 2008.
[6] S. Kajita, T. Yamaura, and A. Kobayashi, “Dynamic Walking Control of a Biped Robot Along a Potential Energy Conserving Orbit,” IEEE Transactions on Robotics and Automation, vol. 8, no. 4, pp. 431–438, 1992.
[7] 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。
[8] N. Motoi, T. Suzuki, and K. Ohnishi, “A Bipedal Locomotion Planning Based on Virtual Linear Inverted Pendulum Mode,” IEEE Transactions on Industrial Electronics, vol. 56, no. 1, pp. 54 – 61, 2009.
[9] 張宏毅，「基於能量函數與倒單擺模型之雙足機器人步態軌跡規劃與控制」，國立台灣大學碩士論文，民國一百年。
[10] J. Morimoto, G. Endo, J. Nakanishi, and G. Cheng, “A Biologically Inspired Biped Locomotion Strategy for Humanoid Robots: Modulation of Sinusoidal Patterns by a Coupled Oscillator Model,” IEEE Transactions on Robotics, vol. 24, no. 1, pp. 185 – 191, 2008.
[11] K. Miura, M. Morisawa, S. Nakaoka, F. Kanehiro, K. Harada, K. Kaneko, and S. Kajita, “Robot Motion Remix based on Motion Capture Data - Towards Human-like Locomotion of Humanoid Robots,” IEEE International Conference on Humanoid Robots, pp. 596 – 603, 2009
[12] S.Wang, W. Chaovalitwongse, R. Babuska, “Machine Learning Algorithms in Bipedal Robot Control,” IEEE Transactions on Systems, Man, and Cybernetics, Part C: Applications and Reviews, vol. 42, pp. 728 – 743, 2012.
[13] J.G. Juang, C.S. Lin, “Gait Synthesis of a Biped Robot Using Backpropagation Through Time Algorith,” IEEE International Conference on Neural Networks , vol. 3, pp. 1710 – 175, 1996
[14] Y.D Hong, C.S Park, and J.H Kim, “Human-Like Stable Bipedal Walking with a Large Stride by the Height Variation of the Center of Mass Using an Evolutionary Optimized Central Pattern Generator,” IEEE International Conference on Industrial Electronic Society, pp. 177 – 182, 2011.
[15] PhaseSpace Motion Capture System information [Online]. Available:
http://www.pitotech.com.tw/
[16] Y.J. Lee, W.F. Hsieh, and C.M. Huang, “ε-SSVR: a Smooth Support Vector Machine for ε-Insensitive Regression,” IEEE Transactions on Knowledge and Data Engineering, vol. 17, no. 5, pp. 678 – 685, 2005.
[17] Robotics and Mechanisms Laboratory information [Online]. Available:
http://www.romela.org/
[18] D.A. Winter, “Foot Trajectory in Human Gait: A Precise and Multifactorial Motor Control Task,” Journal of the American Physical Therapy Association, vol. 72, pp. 45 – 53, 1992.
[19] Mathork Polynomial Curve Fitting [Online]. Available: http://www.mathworks.com/help/matlab/ref/polyfit.html
[20] S.T. John, P.L. Bartlett, “Structural Risk Minimization Over Data-Dependent Hierarchies,” IEEE Transactions on Information Theory, vol. 44, pp. 1926 – 1940, 1998.
[21] C.M. Huang, Y.J. Lee, D.K. J. Lin, and S.Y. Huang, “Model Selection for Support Vector Machines via Uniform Design,” A special issue on Machine Learning and Robust Data Mining of Computational Statistic and Data Analysis, vol. 52, pp335 – 346, 2007.
[22] Laboratory of Movement Analysis and Measurement ： Foot clearance [Online]. Available: http://lmam.epfl.ch/page-73893-en.html