Simulating bipedal walking using a translating center of pressure

During walking, foot orientation and foot placement allow humans to stabilize their gait and to move forward. Consequently the upper body adapts to the ground reaction force (GRF) transmitted through the feet. The foot-ground contact is often modeled as a fixed pivot in bipedal models for analysis of locomotion. The fixed pivot models, however, cannot capture the effect of shift in the pivot point from heel to toe. In this study, we propose a novel bipedal model, called SLIPCOP, which employs a translating center of pressure (COP) in a spring loaded inverted pendulum (SLIP) model. The translating COP has two modes: one with a constant speed of translation and the other as the weighted function of the GRF in the fore aft direction. We use the relation between walking speed and touchdown (TD) angle as well as walking speed and COP speed, from existing literature, to restrict steady state solutions within the human walking domain. We find that with these relations, SLIPCOP provides steady state solutions for very slow to very fast walking speeds unlike SLIP. SLIPCOP for normal to very fast walking speed shows good accuracy in estimating COM amplitude and swing stance ratio. SLIPCOP is able to estimate the distance traveled by the COP during stance with high precision.