When linear control theories are apllied for the control of a nonlinear system, the performance of the control system depends on the accuracy of the linear model obtained by assumption and approximation. But for some nonlinear system it is often difficult to find a proper linear model Tilt-rotor aircraft is such a nonlinear system.
This paper suggests neuro-fuzzy control as a suitable control strategy for the control of a tilt-rotor aircraft. In a fuzzy control system, fuzzy reasoning is performed by a logic operation with a control rule in the knowledge base. The reasoning process is considered as a pattern recognition problem between input lingustic variables and output linguistic variables. In this paper the neural network technique is adopted for the implementation of the fuzzy reasoning. Such a control system is rule-dependent rather than data-dependent as the conventional control systems are. A neuro-fuzzy controller is designed for the tilt-rotor aircraft, and the robustness to the parameter variations and to the changes in operating points is studied.
