▎ 摘 要
NOVELTY - The method involves obtaining a joint activity signal of a subject. The joint activity signal is collected by a graphene flexible sensor. Knee joint angle data is estimated through a curve extreme gradient boosting (CurXGB) model based on the joint activity signal of the subject. The knee joint angle data is determined corresponding to an angle of a video game. The video game is completed by the subject by utilizing joint action. A predicted value is output when input data is greater than a value of an optimal splitting point when prediction process is performed. USE - Graphene flexible sensor based knee joint rehabilitating method. ADVANTAGE - The method enables adopting a CurXGB regression algorithm to form a continuous prediction curve, providing a movement curve combined with human joint movement for a rehabilitation exoskeleton robot, applying characteristics of the graphene flexible sensor to knee joint rehabilitation process, realizing knee joint rehabilitation training process by a knee joint rehabilitation system based on the graphene flexible sensor, and increasing participant degree of the subject. DETAILED DESCRIPTION - An INDEPENDENT CLAIM is included for a graphene flexible sensor based knee joint rehabilitation system. DESCRIPTION OF DRAWING(S) - The drawing shows a block diagram of a graphene flexible sensor based knee joint rehabilitation system. (Drawing includes non-English language text).