A. Hiliuta, R.M. Botez (Canada), and M. Brenner (USA)

Fuzzy set techniques, clustering, grid partition, adaptive neural fuzzy inference system, aerodynamics, least squares

In this paper, the F/A-18 SRA aircraft is modeled by finite element methods by use of STARS software at NASA Dryden Flight Research Center and the aerodynamic forces acting on the F/A-18 SRA are calculated in the frequency domain and need to be approximated in the Laplace domain by the classical Least Square LS method, in order to study the effects of the control laws on the flexible aircraft structure and calculate the flutter velocities and frequencies. We show here that the Least Square approach fails to provide a correct solution in the case where the aerodynamic forces data are not calculated for a certain range of evenly spaced reduced frequencies. We further show that the combination between fuzzy clustering techniques and the shape-preserving approximation may be used to obtain the intermediate values of unsteady aerodynamic forces in between the known reduced frequencies points where these forces are calculated. Is shown that, if the reduced frequencies range is carefully chosen (evenly spaced reduced frequencies), then flutter speeds and frequencies are well calculated by the common classical Least Square method. Finally, we need to be careful in the choice of reduced frequencies range, which might determine the method to be used.

Important Links:

Go Back