透過您的圖書館登入
IP:3.83.81.42
  • 學位論文

輔助型下肢外骨骼機器人控制系統之研究

A Study on the Control System of Assistive Lower Limb Exoskeleton Robot

指導教授 : 林達德

摘要


外骨骼機器人是一種提升人體活動能力的裝置,分成上肢型、下肢型與全身型的外骨骼機器人,而本研究的對象是下肢型外骨骼機器人,應用於協助下肢行動不便的患者以提升日常生活之便利性與日常復健用途。本研究承接上一代外骨骼機器人研究成果 (張,2015),繼續向下針對其控制系統進行改良。上一代外骨骼機器人利用使用者身體參數調整客製化的步態控制外骨骼機器人,搭配胸前的慣性量測單元與有限狀態機來辨識目前的動作種類。然而在這些動作的運作過程,並沒有控制迴路來監控動作的流暢程度,也無法得知使用者目前步行狀態。因此本研究開發了一整套控制流程,包含起始步態意圖,外骨骼機器人能根據步態意圖控制器決定啟動左右腳的順序;步態行走速度控制,可以根據步態行走速度改變行進中的速度;步態事件偵測演算法與步態軌跡產生器。外骨骼機器人在不間斷的最快步行速度可達3 s/step,相當於一分鐘可以走20步,此步速是上一代的4倍,並且透過客製化的角速度步態可以使每個關節在運作的過程中並非都是以等角速度轉動,使外骨骼機器人更加貼近正常人行走之步態,並且藉由即時控制角速度步態軌跡的大小,可以將外骨骼機器人的步速依使用者的意願控制在一定的範圍。在硬體改良的部分,本研究設計並製作了腳踝機構,此機構足以支撐機器人機構之重量,以減低受試者在穿戴時之不適;另外一部分是將原本的運算單元改成嵌入式系統,以提升實驗過程或使用過程之便利度。

並列摘要


Exoskeleton robot is a kind of equipment that can enhance human mobility, and there are three categories including upper limb, lower limb and whole body exoskeleton robot. The purpose of this research is to focus on the lower limb exoskeleton, and wish to apply this equipment on increasing the quality of the patients’ activities of daily living (ADL) as well as helping patient to recover from walking diseases. In this research, the objective is to undertake the former version of the lower limb exoskeleton robot, and modifies its control system so that robot can act more like a real human. In the previous version (Chang, 2015) uses the parameters from the user’s physical dimension to generate customized gait model to control exoskeleton robot, and also used the inertial measurement unit on the chest to determine current finite state. However, in the process of these motion, there is no any feedback control loop to monitor the whole process, and the exoskeleton system is not aware of the current walking status. As a result, this research introduced a whole new version of control system, including gait intension estimator, cadence control, gait phase detection algorithm, and trajectory generator. In this version of exoskeleton is able to walk in the maximum speed of 3 s/step, that is to walk 20 steps within a minute, this result is 4 time faster than the former version. Last but not least, a new ankle-foot orthosis and an embedded system was designed, this enables user to walk more comfortably and to increase mobility of the exoskeleton robot. This version of robot not just walks faster than the former version, but walk more like a real human.

參考文獻


張家瑋。2015。輔助行走下肢外骨骼機器人之設計。碩士論文。台北:台灣大學生物產業機電工程學研究所。
Anam, K., & Al-Jumaily, A. A. 2012. Active exoskeleton control systems: State of the art. Procedia Engineering, 41, 988-994.
Dejnabadi, H., Jolles, B. M., & Aminian, K. 2008. A New Approach for Quantitative Analysis of Inter-Joint Coordination During Gait. IEEE Transactions on Biomedical Engineering, 55(2), 755-764. doi:10.1109/TBME.2007.901034.
Dinh, B. K., Cappello, L., & Masia, L. 2016, 26-29 June 2016. Localized Extreme Learning Machine for online inverse dynamic model estimation in soft wearable exoskeleton. Paper presented at the 2016 6th IEEE International Conference on Biomedical Robotics and Biomechatronics (BioRob).
Dollar, A. M., & Herr, H. 2008. Lower extremity exoskeletons and active orthoses: challenges and state-of-the-art. IEEE Transactions on robotics, 24(1), 144-158.

延伸閱讀