25th Congress of International Council of the Aeronautical Sciences, 3 - 8 September 2006, Hamburg, Germany
Paper ICAS 2006-6.5.3


A NEW APPROACH TO NONLINEAR MODELING OF HIGHLY MANEUVERABLE AIRCRAFT USING NEURAL NETWORKS

F. Saghafi, B. M. Heravi
Sharif Univ. of Technology, Iran

Keywords: Aircraft, Modeling, Simulation, Identification, RNN

In this work, the ability of neural networks in modeling the dynamic behavior of highly maneuverable aircraft is investigated. Six modified Elman networks were trained to model the F-16 dynamics working completely off-line. The obtained network generalization for arbitrary pilot inputs and different maneuvers was found quite acceptable.


view full paper