A Biped Robotic Hybrid Control Based on Fuzzy Neural Networks

Abstract

The paper presents a fuzzy neural networks hybrid control of biped robot control problem. This method integrates the advantages of fuzzy neural network and H ∞ controller and inverse system method. On the one hand, regarding and construction errors of FNN as external disturbance, using H ∞ controller attenuate disturbance to a prescribe level. On the other hand, the strong approximate capability of FNN construct the inverse system and offer efficient system information to H ∞ controller. The stability of close loop system is analyzed, L 2 gain can be attenuated by the presented fuzzy neural network structure and adaptive algorithm. In addition, the partitioned neural networks are applied in robot control problems.