Simulation of human locomotion using a musculoskeletal model

Human gaits in a sagittal plane are simulated throughout a musculoskeletal model with a simple feedback postural controller. Hill-type muscletendon models generate joint motions over each leg. Typical human gait data about joint torques and angles is used to estimate muscular activation signal profiles by a simple optimization technique. A simple feedback controller updates the activation profiles to achieve stable dynamical walking simulation through optimization with a genetic algorithm. Simulation study demonstrates the proposed approach achieves realistic human walking behavior with neurophysiologically plausible muscular activations.