Estimation of Performance Parameters of Turbine Engine Components Using Experimental Data in Parametric Uncertainty Conditions

Abstract

Gas Path Analysis and matching turbine engine models to experimental data are inverse problems of mathematical modelling which are characterized by parametric uncertainty. This results from the fact that the number of measured parameters is significantly lower than the number of components’ performance parameters needed to describe the real engine. In these conditions, even small measurement errors can result in a high variation of results, and obtained efficiency, loss factors etc. can appear out of the physical range. The current methods of engine model identification have developed considerably to provide stable, precise and physically adequate solutions. Presented in this work is an estimation method of engine components’ parameters based on multi-criteria identification which provides stable estimations of parameters and their confidence intervals with the known measurement errors. A priori information about the engine, its parameters and performance is used directly in the regularised identification procedure. The mathematical basis for this approach is the fuzzy sets theory. Forming objective functions and scalar convolutions synthesis of these functions is used to estimate gas-path components’ parameters. A comparison of the proposed approach with traditional methods showed that its main advantage is high stability of estimation in the parametric uncertainty conditions. Regularization reduces scattering, excludes incorrect solutions which do not correspond to a priori assumptions, and also helps to implement the Gas Path Analysis at the limited number of measured parameters. The method can be used for matching thermodynamic models to experimental data, Gas Path Analysis and also adapting dynamic models for the needs of the engine control system.

Aircraft Turbine Engine Automatic Control Based on Adaptive Dynamic Models

“…Therefore, this adaptation direction concentrates on the correction of the static model. With this purpose, correcting parameters are introduced into the model ( 1) and the static model has the following form: (6) We developed an approach to the determination of the parameters in need for correction [11] based on the least Square method. To improve its robustness, this method is modified by combining the traditional function of empirical risk: (7) with the function of a priori risk: (8) where:…”

“…The proposed estimation procedure is shown in Fig. 5 and described in the paper [11] in application to a problem of the engine technical state analysis. It is also applicable for determining the best estimates of parameters for the model (1) correction.…”