Hybrid control of biped robots in the double-support phase via H/sub (infinity)/ approach and fuzzy neural networks

A quadratic stabilisation fuzzy neural network hybrid control scheme is proposed for biped robots in the double-support phase. First, the authors considered the holonomic constraints. The walking locomotion of biped robots in the doublesupport phase is modelled as a reduced order position/force hybrid model, which integrates the position/force hybrid model with the reduced-order model. Then a new fuzzy neural network hybrid control scheme is presented, in which H/sub (infinity)/ approach, inverse system method and variable structure control are used to attenuate the effect of external disturbances and parametric uncertainties. Simulation results are reported in order to show the performance of the proposed control scheme.