부산대학교 도서관

Joint torque estimation techniques are used for evaluating human motor function. Particularly, ankle joint torque is essential for evaluating the walking function. Ankle joint torque can be derived by fixing the metatarsophalangeal (MTP) joint and ankle circumference and measuring force and torque from force sensors placed at the two fixation points. Considering implementation effort, it is desirable to use a 6-axis sensor or a sensor with fewer measurable axes. Therefore, in this study, we compared multiple sensor arrangements and calculation methods in the measurement of ankle joint torque. Through theoretical and empirical evaluations, the conditions under which these methods are effective were clarified. Based on the investigation, we propose a method to arrange a 6-axis force sensor only at the toe. One concern in the theoretical investigation was the error due to the moment caused by the fixation of the heel side. However, in measurement experiments using an ankle model and subjects, the accuracy was almost the same as when two 6-axis force sensors were used. The results indicates that the effect of this error factor is limited: hence, this method is considered to be useful as a method of both accuracy and sensor reduction.