Intelligent landing control based on neural-fuzzy-GA hybrid system

Abstract

This work presents three intelligent aircraft automatic landing controllers that use fuzzy system, hybrid fuzzy-neural system and hybrid fuzzy-GA system to improve the performance of a conventional automatic landing system. In this study a multi-layered fuzzy modeling network is used as the controller. Control gains are selected by a combination method of a nonlinear control design, a neural network, and genetic algorithm. Comparisons on different control schemes are given. Simulation results show that the proposed automatic landing controllers can successfully expand the safety envelope of an aircraft to include severe wind disturbance environments without using the conventional gain scheduling technique.