1 * Dr., Research Associate ‡ Professor Dr., Head of laboratory Copyright ©2009 by the American Institute of Aeronautics and Astronautics Inc. All rights reserved ISABE-2009-1301 Estimation of gas turbines gradual deterioration through a Dempster-Schafer based fusion method C. Romesis*, K.Mathioudakis ‡ Laboratory of Thermal Turbomachines National Technical University of Athens URL:http://www.ltt.mech.ntua.gr/ , Email: kmathiou@central.ntua.gr ABSTRACT This paper presents a fusion procedure of independently acting diagnostic methods, allowing gas turbines health condition assessment given a series of measurements. The proposed procedure incorporates a fusion technique, which is based on the Dempster-Schafer theory. The novel element of the method is its ability to cope with the problem of overall engine gradual performance deterioration, instead of identification of individual component fault events. The effectiveness of the technique is evaluated through its application on scenarios representing drifting gas turbine faults encountered in practice, using independently acting diagnostic methods, already established. Through this application, the efficiency of the proposed fusion procedure is demonstrated, along with the improvement it provides over its constituent methods to both the accuracy and the reliability of diagnosis. NOMENCLATURE ⊕ Masses combination operator (eq.9) ℵ(μ,ı) Normal distribution with mean value (μ) and standard deviation (ı) DM-i i-th diagnostic method (Figure 1) DOD Domestic Object Damage D-S Dempster-Schafer EPM Engine Performance Model f Vector of the health parameters of the engine f act i,j Actual value of f i at point j (eq. 15) f DM-i Health parameters estimations from method DM-i f DS Health parameters estimations from D-S fusion technique f i Health parameter i f i,j Estimated value of f i at point j (eq. 15) f j i Estimated value of f i , at point j (eq. 14) f j i,LS Estimated value of f i , at point j, calculated through a fitted polynomial curve (eq. 14) FOD Foreign Object Damage HPC High Pressure Compressor HPT High Pressure Turbine k Sum of product of masses (eq. 10) and (eq. 13) LPT Low Pressure Turbine M No. of points of f i estimation (eq. 14) m(x) Mass of element x, according to D-S theory m i (x) Mass of x, provided by method DM-i OF Objective Function P(f i ∈φ j ) Probability of f i to lie within the interval of values φ j P(Θ) Power-set of Θ P i (f i ∈φ j ) Probability of f i to lie within the interval of values φ j , provided by method DM-i SEi Efficiency factor at station i of the engine s fi Standard deviation of the estimations of f i (eq.15) SWi Flow factor at station i of the engine Θ Environment, according to D-S theory θ i i-th element of environment Θ ı DM-i Standard deviation of the estimations from method DM-i ı fi Standard deviation of f i estimation (eq. 14) Υ Vector of measured quantities on an engine φ j A interval of values of a health parameter INTRODUCTION Gas turbines condition monitoring and health assessment is vital in view of the benefits associated with its implementation. The field of gas turbines diagnostics has, thus, an important development in the last decades. Numerous diagnostic methods and tools have been proposed to date, all aiming at an assessment of gas turbines health condition from several sources of available information. An important issue that gas turbine researchers are dealing with is the deterioration of performance with time, which results from a number of physical mechanisms. The ability of the user to identify the current condition of a deteriorated engine is a key factor to ensuring reliable operation and optimal usage. The ability to forecast deterioration and the importance of producing models providing such a possibility has been