A new method of fault diagnosis for aeroengines with dispersedly clumped gas path parameters

Abstract

In an aircraft fleet with multiple aeroengines in service, the gas path parameter measurement values from engines of the same type are dispersed within a certain range due to manufacturing technology, assembly tolerances, operation conditions, and maintenance. This dispersion cannot be ignored and makes it impossible to apply unified health diagnosis criteria to all engines. To address this issue, a fault diagnosis solution for aeroengines with dispersedly clumped gas path parameters based on meta aero-engine is proposed. The concept of meta aero-engine is proposed to derive an engine group with dispersedly clumped parameters. To diagnose engines, a sphenophyllum classifier is proposed, which has a unified judgement criterion. A new weak classifier ensemble method based on divide-and-conquer is proposed to enhance the classification ability of the sphenophyllum classifier. The results of the basic classifier are partitioned and then analyzed in combination with the results of other classifiers in the classifier ensemble method. The effectiveness of the proposed method is verified through turbofan engine simulation dataset. The proposed method has strong interpretability for engine fleet with dispersedly clumped gas path parameters.