Multibody Dynamics Integrated With Muscle Models and Space-Time Constraints for Optimization of Lifting Movements

Abstract

A multibody dynamics model integrated with space-time constraints based optimization is presented in this paper for generating optimal trajectories of human lifting movements. “Space-time constraints” is a two-point boundary value dynamic optimization technique developed for animation of computer graphics characters and has a significant potential for biomechanics and other mechanical movement based dynamic optimization problems. Optimization results demonstrate the ability to consider different preferences for minimizing the loading of specific joints such as an ankle, or a knee, or a shoulder during the lifting motion and the resulting lifting trajectories are shown to be different. Lumped muscle models to generate the joint torques are incorporated at five joints to model the actuation effects of the muscular system during the dynamic movement. The dynamic optimization is then based on the muscle activation parameters instead of the traditionally used joint torques. The muscle activation model optimization is shown to correlate better with the actual motion tests conducted by the VICON video capture and test data analysis system.