融合静电信号和气路参数的发动机性能评估方法

A Method for Engine Performance Evaluation by Fusing Electrostatic Signal and Gas Path Parameters

关键词:  航空发动机  静电监测  性能评估  逻辑回归  实验

Key words:  Aero-engine  Electrostatics monitoring  Performance evaluation  Logistic regression  Experiment

摘要:

为了研究静电监测技术对航空发动机的监测效果，开展了航空涡喷发动机静电监测台架实验，获得了140个完整试车阶段的静电信号数据。针对发动机传统气路性能评估方法的信息源有限的问题，提出了一种融合静电信号和气路参数的发动机性能评估流程和方法，进一步采用了基于逻辑回归模型的评估算法对发动机的综合性能进行量化评估，通过融合试车实验的静电数据和性能参数，对发动机的健康状态评价方法进行研究。结果表明:当健康值为0.9以上可以认为发动机为健康状态，当健康值小于0.3时可以认为发动机已经发生了较为严重的退化，所提出的融合评估方法要优于传统的性能参数评估方法，能够提前预警气路性能严重退化。

Abstract:

In order to study the effect of electrostatic monitoring technology on aero-engine， an electrostatic monitoring bench experiment is carried out on turbojet engine， and the electrostatic signal data of 140 complete testing periods are acquired. In view of the limited information source of traditional gas path performance evaluation method， a method of engine performance evaluation based on electrostatic signal and gas path parameters is proposed. The evaluation algorithm based on logistic regression model is further used to quantify the comprehensive performance of the engine evaluation， through the integration of static test data and performance test parameters， a comprehensive evaluation of the health status of the engine. The results of data analysis show that when the health value is more than 0.9， the engine can be considered healthy， and when the health value is less than 0.3， the engine can be considered to have undergone severe degradation， the proposed fusion evaluation method is superior to the traditional method of evaluating performance parameters， and can early warn the serious deterioration of the gas path performance and provide maintenance windows for maintenance personnel.