Vitale, Antonio (2013) Multi-Step Estimation Approach for Aerospace Vehicle System Identification from Flight Data. [Tesi di dottorato]

Anteprima

Testo PhD-dissertation_VITALE.pdf Download (7MB) | Anteprima

Abstract

System identification from flight data is an essential task in aerospace field, both for research and industrial activities. Indeed, ground based tests are not fully exhaustive of the vehicle behaviour and in-flight experimentation is often mandatory. In particular, there is a specific interest in obtaining vehicle model characteristics from flight data, in order to better understand theoretical predictions of physical phenomena, to validate wind-tunnel test results and to get more accurate and reliable mathematical models of the vehicle. The availability of these models is one of the critical items in order to guarantee the competitiveness of the aerospace industry, because it allows designing flight control law, evaluating vehicle performance and handling qualities, performing fault diagnosis and reconfiguration, developing high fidelity simulators, while reducing the flight test time and therefore reducing cost, risks and time to market of new products. Although in the last decades several methodologies have been developed and many applications have been successful demonstrated, there are still open problems and challenges in system identification, mainly related to model complexity, high bandwidth requirements, constraints on flight test manoeuvres due to safety reason, dynamically unstable response, accurate characterisation of model uncertainties. In the present work an innovative system identification methodology is described, which is suitable for dealing with some of the above listed challenges. The proposed methodology is implemented in the framework of a multi-step approach, which decomposes the complex starting identification problem in simplified sub-problems and allows specifying a suitable estimation technique compliant with each sub-problem objective, exploiting the advantage of both time-domain and frequency-domain methods. The straightforward combination of several estimation techniques brings to an identified model which is applicable in a wide frequency range. Furthermore, the proposed methodology is suitable to deal with problems where identification manoeuvres are minimised, indeed the identification can be executed only for the sub-model which is in fact identifiable. Another relevant peculiarity of the proposed approach concerns the exploitation of all the available a priori information and the rigorous management of all the uncertainties involved in the system identification procedure. As a result, a reliable, complete, and structured statistical characterisation of the identified model could be obtained. Proposed methodology is applied in this thesis to determine the dynamical characteristics of rotorcraft vehicles and the transonic aerodynamic model of an atmospheric re-entry space demonstrator. Its effectiveness is demonstrated through numerical assessments, which enhanced the capability to catch the true values of the model parameters and to reproduce the phenomena of interest. Moreover, the application to actual flight data of the CIRA FTB1 re-entry demonstrator allowed to validate and refine the available pre-flight aerodynamic model of the vehicle, in terms of nominal values update and significant reduction on model uncertainties. These results justifies the importance of flight tests and, in particular, of system identification from flight data. The availability of an updated aerodynamic model represents a fundamental step for the development of the upgraded version of the Guidance, Navigation and Control system for the next missions of the same configuration, where the accuracy of estimates and the reliability of the model over an expanded flight envelope will be carefully analysed and assessed. All the activities hereafter reported have been basically performed at the Italian Aerospace Research Centre (CIRA), where the author is the scientific coordinator of the Modelling and Simulation Laboratory, with the collaboration, advice and support of University “Federico II” that hosted the PhD period.

Downloads

Actions (login required)

Modifica documento