2021 年 12 月
第 42 卷 第 12 期
推 进 技 术
JOURNA L O F PRO PU L S ION TECHNO LOGY
Dec. 2021
Vol.42 No.12
基于试飞数据的航空发动机状态监测与故障诊断 *
潘鹏飞
(中国飞行试验研究院,陕西 西安 710089)
摘 要:为了实时监控航空发动机工作参数变化情况,快速及时地预测并诊断发动机故障,本文基
于实际试飞数据建立了航空发动机ANN-NARX 参数预测模型,考虑到建模样本量大、模型结构复杂、
训练时间长、输入输出延迟等因素,采用遗传算法对模型的最小数据样本需求和结构进行了改进优化,
并利用蒙特卡洛方法确立了参数预测模型的自适应告警门限,同时,基于构建奇偶空间残差模型实现了
航空发动机典型故障诊断。结果表明:实际试飞中只需有限架次试飞数据的训练学习,即可得到发动机
参数预测模型,高压转子转速、压气机出口压力、低压涡轮出口温度及滑油回油温度相对误差最大值分
别为1.0%,1.7%,0.2%和1.2%,综合模型建模误差和参数测量误差后的自适应告警门限有效降低了模
型预测结果的不确定性,在已有数据样本集上的典型故障识别率达到95.2%。
关键词:航空发动机;飞行试验;状态监测;神经网络;故障诊断
中图分类号:V231.3 文献标识码:A 文章编号:1001-4055(2021)12-2826-12
DOI:10.13675/j.cnki. tjjs. 200707
Flight Data Based Condition Monitoring and Fault
Diagnosis of Aero-Engine
PAN Peng-fei
(Chinese Flight Test Establishment,Xi’an 710089,China)
Abstract:During flying test life cycle,aircraft engine conditions change greatly and faults have been en⁃
countered frequently. There always exist urgent needs about monitoring parameters trending on-line,predicting
possible faulty condition and diagnosing the specific type when faulty condition encountered. The problem of con⁃
dition monitoring and fault diagnosis based on flight test data has been studied in this paper. ANN-NARX param⁃
eters predicting model of aero engines has been built based on actual flight test data. Considering large demands
on data samples,the complex and large design space of ANN model,consequently long training time and inputoutput time delaying,the model architectures and minimum sample demands have been optimized based on
evolving algorithms. The self-adapting thresholds of predicting model have been set using Monte-Carlo method.
The specific fault diagnosis has been realized by constructing parity space residual model. All models in this pa⁃
per have been tested through flight data and applied in actual flying test. The monitoring model could be built
based on limited flights in actual flying test. The maximum relative error of high-pressure spool speed,pressure
in compressor outlet,total temperature in low-pressure turbine outlet and temperature of all returned oil is
1.0%,1.7%,0.2% and 1.2%,respectively. The model predicting uncertainty could be greatly reduced using
adaptive thresholds by considering both modeling error and measurement uncertainty. The ratio of detecting and
* 收稿日期:2020-09-11;修订日期:2020-12-10。
通讯作者:潘鹏飞,硕士,高级工程师,研究领域为动力装置工作特性与性能特性试飞。
引用格式:潘鹏飞 . 基于试飞数据的航空发动机状态监测与故障诊断[J]. 推进技术,2021,42(12):2826-2837. (PAN
Peng-fei. Flight Data Based Condition Monitoring and Fault Diagnosis of Aero-Engine[J]. Journal of Propulsion
Technology,2021,42(12):2826-2837.)