Identification of dynamic characteristics of two-shaft turbofan engine

Abstract

Mathematical models are an important tool for designing engines and their automatic control systems. The models are used in the simulation modeling of the control object, in the analysis, synthesis, semi-realistic modeling, and in the synthesis of the engine's automatic control algorithms. The accuracy of a model is important and cannot be ensured without individual adjustments to a specific engine. The development direction associated with the introduction of adaptive automatic control systems (ACS) is promising. The key feature of these systems is matching algorithms and the mathematical model's self-adjustment to the engine's real characteristics. This paper considers the formation of models using the method of least squares (LSM) based on the measured parameters of the engine's working process. An urgent need is to ensure sufficient information to achieve a given level of accuracy. This check must be performed a priori (to predict the composition of engine operating modes and the volume of registered information) and a posteriori. The difficulty in evaluating the dynamic characteristics of a two-shaft engine lies in its larger number than a gas generator. The formation of a model of the parameters of a two-shaft engine can be done in many ways. In this work, such a model of a two-shaft engine is used, the parameters of which reflect the engine's dynamic properties and the dynamics of influences and separately characterize the oscillations. The subject of the study is the process of forming a dynamic mathematical model (MM) of a two-shaft gas turbine engine using real data for the subsequent application of this model for solving problems of control and diagnostics of onboard systems. The goal is to determine the dependence of the errors of estimation of the dynamic parameters of mathematical models on the influencing factors in the conditions of real change of these factors over time. Tasks considered in the work: formation of a dynamic model of a two-shaft engine and reduction of the number of independent dynamic characteristics, formation of the LSM-functional for the evaluation task, determination of the evaluation errors of dynamic coefficients; analysis of influencing factors, as well as determination of dependencies between factors and errors. For this purpose, the methods of linear dynamic systems theory and statistical evaluation are used. The following results were obtained: a ratio was determined for calculating the error in estimating the dynamic characteristics of a two-shaft engine when the fuel consumption changes in leaps and bounds. It was studied on the example of a real engine, and the regularities between the dynamic characteristics, which simplify the estimation of parameters, were revealed. Scientific and practical novelty: for the first time, a ratio was obtained that determines the estimation errors of time constants based on the specified values of measurement errors, the amplitude of fuel consumption, the frequency and duration of observation, as well as the dynamic characteristics of the system. These relations are given in dimensionless coordinates and can be applied to any twin-shaft turbojet engine when planning experiments and debugging on-board algorithms for self-tuning models.