FLIGHT CONTROL SYSTEM DESIGN FOR COMMERCIAL AIRCRAFT USING NEURAL NETWORKS

Abstract

This paper describes a neural network based approach for flight control design of commercial aircraft. A neural flight control system is used to actuate the gains of a linear controller for gain scheduling purposes without any explicit system identification. The neural network system uses reference models to specify desired handling qualities. Simulation results demonstrate the good tracking performance of the proposed control approach and the learning capability of the proposed neural network controller by showing rapid convergence of the network weights.