A Fuzzy Logic Approach to Gas Path Diagnostics in Aero-engines

Abstract

Engine-related costs contribute a large fraction of the direct operating costs (DOCs) of an aircraft, because the propulsion system requires a significant part of the overall maintenance effort. Thus, to ensure competitive advantage in the aeroengine market, health monitoring systems with gas path diagnostics capability are highly desirable.In this chapter, an application of fuzzy logic technology to gas path diagnostics for aero-engines performance analysis is presented and the setup procedure for a modern civil turbofan is described, as an example. The objective is to estimate the changes in engine component performance due to the engine degradation over time from the knowledge of only a few measurable parameters, inevitably affected by noise. This is a novel process that achieves effective diagnosis by means of a rule-based pattern-recognition methodology founded on fuzzy algebra, developed to provide an alternative technology versus conventional estimation algorithms.The inherent capability of fuzzy logic to deal with gas path diagnostics difficulties, thanks to the use of fuzzy set theory and its rule-based nature, is highlighted. First, the problem of noisy measurements is treated at a fuzzy-set level. Second, at the system level the definition of fuzzy rules is used to map input sets of measurements into output faulty classes of performance parameters in a constrained search space; this enables a problem reduction aimed at overcoming the fact that the analytical formulation is undetermined.The process quantifies the performance parameters' deteriorations through a nonlinear approach, even in the presence of noisy measurements that typically complicate the diagnostic assessment. The diagnostics model's setup as well as its outcome can be attained in a relatively short time, making this technique suitable for on-board use. The accuracy of the technique relative to simulated turbofan data is tested and its advantages and limitations are discussed. IntroductionThe performance of an aero-engine deteriorates over time as a consequence of its components' degradation. The identification of the exact component(s) responsible for the performance loss facilitates the choice of the recovery action to be undertaken. An engine gas-path diagnostic process calculates changes in the magnitude of the component performance parameters (e.g., efficiency and flow capacity) given a set of measurements (e.g., temperatures, pressures, shaft speed and fuel flow) through the engine. However, accurate assessment is complicated by

Multiple-Point Adaptive Performance Simulation Tuned to Aeroengine Test-Bed Data

“…The offdesign adapted models were obtained by scaling the compressor maps of the three compressors to minimize the objective functions in Eqs. (2) and (19). The values of the OF for the adapted models used in this work are listed in the last row of Table 5.…”

“…The values of the OF for the adapted models used in this work are listed in the last row of Table 5. Equation (2) represents the sum of the percentage deviations between the models and the data for all 8 measurable parameters. In Table 5, these equations were also used to calculate the error summation for each model.…”

“…Equally of interest to both engine manufacturers and operators is the fact that these deviations between the simulation and the operating engine, if not reduced to an acceptable minimum, can result in misleading gas-path diagnostics (GPD) analyses [1,2]. GPD is used in health monitoring to identify the component(s) responsible for performance deterioration and to quantify the deviations of each component performance parameters (typically efficiencies and flow capacities of compressors and turbines) from a baseline condition represented by the simulation model.…”

Identification of Fuzzy Measures for Machinery Fault Diagnosis

“…In this paper, the fuzzy measure based on partial correlation coefficients is defined. This fuzzy measure is the extension of the fuzzy measure defined by (Taya and Murofushi, 2006). In the definition,…”

“…Several diagnostic systems for rotating machineries have been developed to satisfy this need (Liu et al, 2007). Some of them use fuzzy measures and fuzzy integrals to encompass the existing knowledge of skilled engineers (Marinai and Singh, 2006). However, they still have several problems, such as difficulty in isolating faults generating similar vibration spectra.…”