Investigations for the dynamical analysis of human motion

In this paper, an interdisciplinary combination of several state-of-the-art techniques from multibody dynamics, biomechanics, muscle modeling, optimal control, and visualization for the investigation of human motions is described. The goal is to use (wherever possible) and modify (wherever necessary) existing techniques and to investigate what has to be observed to make them work together efficiently. A separate investigation and implementation of the topics often is not sufficient for a combined approach, e.g., when some very sophisticated muscle models described in the literature yield infeasible computation times, even on modern supercomputers, if they are combined with optimal control techniques. First, some basic elements of multibody dynamics and muscle models are described. Then, a state space model for the complete system is derived and optimal muscle stimulations are computed using optimal control techniques. Finally, extensions for the visualization of biomechanical effects are presented and the application to human kick and jump motions is shown. However, the main focus of the paper is the presentation of the interdisciplinary approach and its cooperating methods and not the in-depth description of individual topics or the investigation of a specific biomechanical application.