This paper proposes a knee joint moment estimating algorithm which utilizes the surface EMGs of involved muscles and joint angle kinematic data. This algorithm will be essential part for the control system design of exoskeletal robotic devices. In order for this algorithm to accurately predict joint moments, it is necessary to know the one’s musculo-skeletal properties, which is virtually impossible. An optimization process is used to determine one’s musculo-tendon characteristics. The proposed method is evaluated through the comparison with the isokinetic experimental results. Based on the analysis on this comparison, it is claimed that the gravity compensation data of the dynamometer (CON-TREX MJ system) contains not only purely gravity compensated portion but also the knee joint moment contributed by the passive muscle fiber forces.

Категория: Ищем научные статьи | Нет комментариев »

Комментарии