Identification of flexible wing aircraft models using hyperbolic metrics

Abstract

This paper proposes a novel discrete time identification method for flexible wing aircraft models from simulated data. The special properties of the dynamics arise from the fact that all the poles are located very close to the unit circle and their location changes with the flight velocity. Using identification criteria based on Euclidean metrics suffers from the problem of obtaining instability in the search and also on the possible high number of poles. The paper offers an alternative approach and associated identification method to obtain the poles directly using Laguerre-representation of the impulse responses and hyperbolic metrics for identification criteria. The method utilises the hyperbolic geometrical properties of the descriptions of discrete-time signals and systems on the unit disc, and is strictly connected to the representations in rational orthogonal bases. Beyond the conceptual clarity, examples also confirm that it forms an adequate basis for estimating poles of systems as a significant part of an identification process.