Skip to Main Content
System Identification is a key technology for the development and integration of modern engineering systems including gas turbines. These systems are highly parametric with complex dynamics and nonlinearities. Small turbojet engines are special class of gas turbines that are suitable for scientific purposes and researches in the area of stability, performance, simulation and fuel control design. The motivation behind the presented study is to improve the speed, quality and cost of engine testing and to gain insight into alternatives to traditional identification methods for gas turbines. It discusses the identification of small turbojet engine dynamics by validating the thermodynamic model of the engine as well as presenting transfer function model, including thrust, shaft speed and turbine exit temperature in relation to the fuel flow. Model identification approach, presented in this paper, is the first to consider frequency-sweep signals, to excite adequately the engine dynamics, as well as to utilize windowing and smoothing techniques, to reduce random errors in the spectral estimates. Frequency sweep provides a fairly uniform spectral excitation over the frequency range of interest. In addition, Chirp-z transform warrants the exact determination of frequency responses, which is robust to uncertainties.
Date of Conference: 18-21 Sept. 2012