23rd Congress of International Council of the Aeronautical Sciences, 8-13 September, 2002, Toronto, Canada
Paper ICAS 2002-4.4.3

AN EXPERT DATA MINING SYSTEM FOR NONLINEAR AEROELASTIC RESPONSE PREDICTION

Y. S. Wong, O. Voitcu, C. A. Popescu
University of Alberta, Canada

Keywords: nonlinear aeroelasticity, limit cycle oscillations, long-term prediction, neural networks, non-linear time series models, unscented filter, data mining, expert system

An expert data mining system (EDMS) is proposed to investigate the nonlinear dynamics in aeroelasticity. A limited data set representing the transient information is provided as input to EDMS, and the output predicts the long-term dynamic behavior of a nonlinear aeroelastic system. The core components in the EDMS include artificial neural networks, nonlinear time series models and the unscented filter.

view full paper