Hopping system control with an approximated dynamics model and upper-body motion

Abstract

A hopping system is highly non-linear due to the nature of its dynamics, which has alternating phases in a cycle, flight and stance phases and related transitions. Every control method that stabilizes the hopping system satisfies the Poincar, stability condition. At the Poincar, section, a hopping system cycle is considered as discrete sectional data set. By controlling the sectional data in a discrete control form, we can generate a stable hopping cycle. We utilize phase-mapping matrices to build a Poincar, return map by approximating the dynamics of the hopping system with SLIP model. We can generate various Poincar, stable gait patterns with the approximated discrete control form which uses upper-body motions as inputs.