Study of Sliding-mode Adaptive Fuzzy Neural Network Control for Controlled Active Motion Apparatus

• Main Authors: Jen-hsien Lo, 羅仁賢
• Other Authors: Tien-chi Chen
• Format: Others
• Language: en_US
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/00645940093971864587

碩士 === 國立成功大學 === 工程科學系碩博士班 === 95 === The Controlled Active Motion (CAM) is the postoperative treatment that is designed to aid recovery after joint surgery. It is thought as a good treatment to accelerate the recovery time for the patient who has had a surgery in clinic. It is shown that the preoperatively existing, proprioceptive deficit can be reduced significantly by combining anterior crucial ligament plasty and neuromuscular training immediately postoperative using the CAM device. The traditional CAM was designed without electronic control such as Camoped (Camoped, Germany). If the patient can not exert too large force just following surgery, the CAM would not be too heavy to be driven. On the contrary, the patient under good recovery condition may want to step on heavier machine. Therefore the resistance of CAM should be designed adjustable for different users. It is easier to adjust the resistance of electronic motorized CAM than in the practical mechanism. In the thesis, the controller is designed to make the mechanism simulate a specified m-b-k system. The parameters of the mechanism vary very sharply due to the foot on the mechanism. Therefore the control scheme of sliding-mode adaptive Fuzzy Neural Network (FNN) was designed. The proposed control scheme combines the merits of sliding-mode and adaptive FNN controller. The simulation and experiment results will be shown to verify the proposed controller.