부산대학교 도서관

A cycling wheelchair (CWC) has been focused as a new rehabilitation tool for hemiplegic patients. These individuals can drive a CWC by rotating the pedals with their non-paralyzed feet. In previous studies, our group has developed a system that allows patients to safely practice driving a CWC in a virtual environment. To confirm the efficacy of rehabilitation with the system, changes in the motor performance of patients driving the CWC should be analyzed. Leg torques of patients rotating the CWC pedals are useful information to evaluate their motor performances. In this study, a new method to estimate the torque of each leg separately was proposed. The proposed method estimates the angle of CWC pedals using the Kalman filter; and, the influence of the weight of users' legs is eliminated based on a human joint model. The accuracy of the proposed method was tested based on an experiment.