Decoupling design of rolling aircraft using adaptive neuro-fuzzy inference systems based on T-S model

Abstract

Considering the coupling problem which is adverse to the control performance of rolling aircraft system, a dynamic decoupling method is proposed combining the feedforward compensation decoupling method and the adaptive neuro-fuzzy inference system. First, the decoupling compensation principle is deduced under the feedforward decoupling framework. And then, the adaptive neuro-fuzzy inference system based on Takagi-Sugeno model is adopted to approximate the decoupling compensation principle. At last, the trained adaptive neuro-fuzzy inference system is used as the dynamic decoupling compensator in series form. A design example of actuator link in rolling aircraft is employed to verify the validity of this method. The simulation results indicate that this method can obtain good decoupling performance in both dynamic and steady responses when it is incapable to get the accurate system model.