Three-dimensional sit-to-stand motion prediction



Journal Title

Journal ISSN

Volume Title



Sit-to-stand (STS) motion is one of the most demanding mechanical tasks in daily life and can be defined as one of the key determinants of functional independence, especially for senior citizens. The mechanics of an STS maneuver was studied extensively in literature, mostly through experimentation. Experimental studies tend to be tedious and time consuming. In contrast to the literature on numerous experimental studies of STS, there are limited studies that were carried out through simulations. In this dissertation, STS tasks are predicted utilizing predictive dynamics. The prediction formulation is constructed as a nonlinear optimization problem. The digital human model has 21 degrees-of-freedom (DOF) for the prediction of the unassisted STS tasks and 30 DOFs for the assisted STS tasks. For the assisted STS task prediction, virtual-individuals are considered to receive assistance from a unilateral grab-rail bar attached on the right side of their bodies. The quartic B-spline interpolation is implemented into the prediction formulation for numerical discretization. A recursive Lagrangian dynamics approach based on the well-established robotic formulation of the Denavit-Hartenberg transformation matrices is utilized for the dynamics of the problem due to its known computational efficiency. Unassisted STS tasks are predicted for healthy young and elderly virtual-individuals, where assisted STS tasks are predicted only for elderly virtual-individuals within three different cases. The first case assumes the elderly virtual-individuals as healthy seniors. In the second and third cases, a knee injury, either in the right or left knee, is implemented into the problem. From the prediction of both the assisted and unassisted STS tasks, kinematics and kinetics of the virtual-individuals are recovered and some insights are obtained.



Biomechanics, Digital human modeling, Sit-to-stand, Optimization, Injury