Online approximators for the actuators FDI of a small commercial aircraft

Authors:	Paviglianiti Gaetano, University of Reggio Calabria, Italy Mattei Massimiliano, University of Reggio Calabria, Italy
Topic:	7.3 Aerospace
Session:	Special Problems of Aircraft Control
Keywords:	Fault detection and isolation, Fault approximation, Aircraft control, Nonlinear adaptive estimators, Radial base function networks.

Abstract

In this paper the problem of detecting and isolating faults on the actuators of a small commercial aircraft using online approximators is dealt with. In particular the capabilities of radial basis neural networks to approximate faults on the engines and elevators of an AP-68 Vulcanair commercial aircraft are investigated also in the presence of model uncertainties and external disturbances. The structure of the aircraft and of the fault models appears to be well suited for this kind of application. Although some effort has to be spent in tuning a certain number of parameters, the numerical results obtained on a full onlinear Matlab-Simulink simulator of the aircraft seem to be encouraging.