Comments on system identification of an UAS model using a subspace method

Abstract

The objective of this study is to present an experimental procedure for identifying the dynamics of an unmanned aerial system (UAS) with a fixed flexible wing. This procedure employs subspace identification techniques, which are particularly suited to the analysis of dynamic systems. In order to comprehend the behaviour of aerodynamic and flight control systems and establish a feedback loop that may be employed to mitigate the impact of structural flexibility, it is imperative to possess a reliable model. The objective of this research is to identify a parametric model for a flexible aircraft from open-loop experimental data by applying the DSRe algorithm. A flight test campaign was conducted using the EOLO, a single-engine aircraft with a wingspan of 4 m and a total weight of 8.87 kg. First, the results of the identification process using synthetic data are presented. The preliminary estimated parameters based on the Ground Vibration Test (GVT) were found to be useful for validating the identified model. Subsequently, the experimental results obtained in open-loop operation demonstrate that subspace algorithms are capable of estimating a suitable state-space model that encompasses the entire frequency range present in the experimental data. It is crucial to emphasise that a significant challenge in developing a representative model for the desired frequency range from the collected data is the necessity for a persistently exciting condition for the input signals.