抄録
This paper proposes an application to estimate the subject-specific geometric and inertial parameters for musculoskeletal model with using time series of motion data. The geometric parameters are identified by solving an inverse kinematics problem with a lower DOF human model introducing virtual generalized coordinates which are constant all time sequences. Base parameters of the geometrically identified model are calculated by real-time identification method using the combination of the motion and force data. All inertial parameters of the subject are estimated based on the base parameters and inertial parameters given by human skin polygonal model. This application provides a convenient means for creating subject-specific musculoskeletal model. The method was tested to identify the parameters of the 155 DOF human musculoskeletal model using a motion capture data and the results are to be shown.
