Predictive Control Model to Simulate Humanoid Gait

Abstract

This article reveals a new approach to model a humanoid gait consist of 5 and 7 links and studying the influence of feet on the overall gait dynamics. Estimated trajectories of limbs have been planned systematically based on equation of motion and their following interpretation for the human movements from their joints and muscles. The human motion is controlled by the central nervous system (CNS) based on model predictive control (MPC). In our projected representations, MPC controller analyses the essential moments at the joints, and these ideal moments are applied to the muscles. furthermore, MPC controller acts the role of the spinal cord in the humanoid CNS. The outcomes of simulation are compared with several examples of real humanoid gait, gained from motion captured systems. According to comparison, the possibility of additional use of the model for individual identification and acknowledgement of gait eccentricities are predictable.