Aero Engine Baseline Modeling Of Gas Path Parameters And State Identification Of Gas Path Based On Flight Data

Aero-engine is the core component that provides power during the flight of the aircraft,and the safety of flight can be affected directly by the state of the aero-engine.Aircraft engine health management system is one of the most effective methods to realize the real-time monitoring and fault diagnosis of the aero-engine.Meanwhile,the engine maintenance cost can be reduced and the economic benefit of airlines can be improved by maintenance decisions through the data support from the aircraft engine health management system.In this paper,the engine post-flight data provided by the airline were used to carry out the following work based on the data processing,baseline modeling and state identification:(1)The outliers and missing values of post-flight QAR data and ACARS data are replaced and filled to improve the overall data quality.Establish the rule of extracting the stable point of QAR data,and the QAR data is synchronized with the ACARS message to obtain higher quality and more comprehensive data.(2)Similar transformation parameters between the inherent laws can be found through the data.The baseline of EGT、FF and N2 can be established by using a multiple linear regression approach based on the data from an aero-engine.The optimal fitting with the grid search method was used in order to improve the precision of baseline data and validated by using several same engines.The modeling precision is set as follows:the error of △FF is within±1%,the error of △N2 is within ±0.2%,and the error of △EGT is within±0.5%.(3)According to the washing time in the maintenance record,the running state of the aero-engine is divided into four categories.The aero-engine state identification model is obtained by using the imaged transformation of the whole flight segment data and combining it with the convolutional neural network.The classification accuracy of this model is over 97%.The precision of the state identification model obtained by only analyzing the cruise data in the same method was lower than that analyzed by using the whole flight segment data.Consequently,reasonable use and extraction of the characteristics of the whole flight segment data are beneficial to obtain the conclusion of the high-precision aero-engine data analysis model.