本論文提出一個多變數強韌控制架構,以控制主動式步態訓練器之直流馬達,產生預設之步態軌跡。本論文使用的主動式步態訓練器,是一種結合連桿機構與馬達之醫療輔具,可以幫助不良於行的病人進行步態訓練。首先,我們藉由系統識別的方式,建立步態訓練器的轉移函數。其次,則回授馬達之電流與轉動位置,設計多變數強韌控制器調變馬達電壓而控制轉速,使步態訓練器產生預定之步態軌跡;因為強韌控制理論可以有效地處理系統不確定性與擾動,故可以提升步態訓練器的性能及強韌性。最後,則將設計之多變數強韌控制器分別安裝於cRIOTM機台與單晶片系統進行實驗。從實驗結果得知,多變數強韌控制,確實可以提升步態訓練器的性能與穩定性,達成步態訓練之目標。
This thesis proposed multivariable robust control for an active gait trainer, which controls the DC motors of a gait trainer to produce the preferred gait traces. The active gait trainer is a medical assistive device, which consists of linkage mechanisms and motors to produce desired gait traces for people with walking disability. First, we obtain the transfer function matrices of the system are by identification techniques. Second, we feedback the tracking errors and motor currents, and design controllers to regulate the motor voltages. In addition, robust control algorithms are applied to deal with system uncertainties and disturbances. Finally, the designed multivariable robust controllers are implemented on cRIOTM and a micro-chip system for experimental verification. From the experimental results, the multivariable robust controllers are shown to effectively improve the stability and performance of the active gait trainer for people with walking disability.