Resolved Multiple Viscoelasticity Control for a Humanoid

This study presents a resolved viscoelasticity control (RVC) method to handle multiple viscoelasticity in a humanoid system. A previous study proposed an RVC by explicitly considering structure-variability and demonstrated biped walking motions on an uneven terrain and a slope. In this study, the RVC is extended to multiple viscoelasticity, namely viscoelasticity on multiple feature points and multiple values in single viscoelasticity. The resolved multiple viscoelasticity control can improve both tracking performance and stability. The results indicate that resolving center of gravity (COG) and swing foot viscoelasticity improves the walking performance and that switching multiple values of the COG viscoelasticity improves the balancing performance. The proposed method is validated by using forward dynamics simulations.