Optimal Control and Forward Dynamics of Human Periodic Motions Using Fourier Series for Muscle Excitation Patterns

Abstract

Forward dynamic simulations of a periodic forearm motion were developed in order to explore the efficiency of using a Fourierseriesbased parameterization function for muscle excitations within dynamic optimization. The specific objectives of this study were to develop such a simulation and validate the predictions. Several timeintegral objective functions, including muscle activation effort and metabolic energy, were used to see the effects of each on the optimal results. For validation, the motion and muscle electromyograms (EMGs) of three adult subjects were captured, where each trial was replicated twice. Fourierseries pattern parameterization was found to be an efficient choice for the muscle excitations in simulating human musculoskeletal dynamics.