基于GA-AANN神经网络的SDQ算法的航空发动机传感器数据预处理

推进技术 ›› 2018, Vol. 39 ›› Issue (5): 1142-1150.

基于GA-AANN神经网络的SDQ算法的航空发动机传感器数据预处理

吕升,郭迎清,孙浩

西北工业大学动力与能源学院，陕西西安 710129,西北工业大学动力与能源学院，陕西西安 710129,西北工业大学动力与能源学院，陕西西安 710129

发布日期:2021-08-15
作者简介:吕升，男，硕士生，研究领域为航空发动机智能控制与健康管理技术。

Aero-Engine Sensor Data Preprocessing Based on SDQ Algorithm of GA-AANN Neural Network

School of Power and Energy，Northwestern Polytechnical University，Xi’an 710129，China,School of Power and Energy，Northwestern Polytechnical University，Xi’an 710129，China and School of Power and Energy，Northwestern Polytechnical University，Xi’an 710129，China

Published:2021-08-15

摘要/Abstract

摘要： 为实现对输入健康管理系统的航空发动机传感器数据进行数据鉴定、故障诊断以及去除噪声信号干扰，提出了一种航空发动机传感器数据预处理方法。针对双通道传感器航空涡扇发动机，搭建了以合理性检验模块和解析冗余检验模块为主要内容的SDQ算法模型，利用遗传算法优化的AANN神经网络实现传感器的解析冗余检验。采用蒙特卡罗仿真方法，将改进的SDQ算法与一种基于最小二乘法的SDQ算法进行对比仿真验证。结果表明，本文提出的SDQ算法在发动机稳态条件下对阶跃故障和漂移故障隔离的平均正确率分别提高了1.7%和19.1%，在发动机动态条件下对阶跃故障和漂移故障隔离的平均正确率分别提高了12.5%和33.8%。且在多传感器故障诊断和除噪方面性能优异，处理后的传感器信号平均信噪比提高了8.27dB。

关键词: 航空发动机传感器；故障诊断；SDQ算法；遗传算法；AANN神经网络

Abstract: In order to realize the data identification， fault diagnosis and noise interference of the aero-engine sensor data of the input health management system， a method of aero-engine sensor data preprocessing was proposed. Aiming at the air turbofan engine of dual-channel sensor， a SDQ algorithm model with reasonableness checks module and analytical redundancy checks module as the main content was established， and the AANN neural network optimized by genetic algorithm was used to realize the analytical redundancy checks of the sensor. The improved SDQ algorithm was compared with a SDQ algorithm based on the least squares method using the Monte Carlo simulation method. The simulation results prove that the average correct rate of step fault and drift fault isolation of the improved SDQ algorithm increased by 1.7% and 19.1% respectively when the engine in steady states， and the average correct rate of step fault and drift fault isolation of the improved SDQ algorithm increased by 12.5% and 33.8% respectively when the engine in dynamic states. The algorithm also has excellent performance in multi-sensor fault diagnosis and noise reduction， and the average signal-to-noise ratio of the processed sensor signal increased by 8.27dB.

Key words: Aero-engine sensor；Fault diagnosis；SDQ algorithm；Genetic algorithm；AANN neural network

吕升,郭迎清,孙浩. 基于GA-AANN神经网络的SDQ算法的航空发动机传感器数据预处理[J]. 推进技术, 2018, 39(5): 1142-1150.

LV Sheng,GUO Ying-qing,SUN Hao. Aero-Engine Sensor Data Preprocessing Based on SDQ Algorithm of GA-AANN Neural Network[J]. Journal of Propulsion Technology, 2018, 39(5): 1142-1150.

参考文献

[1] 龚志飞, 郭迎清. 基于主元分析法的航空发动机传感器故障诊断研究[J]. 计算机测量与控制, 2012, 20(8).
[2] 叶志锋, 孙健国. 应用神经网络诊断航空发动机气路故障的前景[J]. 推进技术, 2002, 23(1): 1-4. (YE Zhi-feng, SUN Jian-guo. Prospect for Neural Networks Used Aeroengine Fault Diagnosis Technology[J]. Journal of Propulsion Technology, 2002, 23(1): 1-4.)
[3] 张书刚, 郭迎清, 陈小磊. 航空发动机故障诊断系统性能评价与仿真验证[J]. 推进技术, 2013, 34(8): 1121-1127. (ZHANG Shu-gang, GUO Ying-qing, CHEN Xiao-lei. Performance Evaluation and Simulation Validation of Fault Diagnosis System for Aircraft Engine[J]. Journal of Propulsion Technology, 2013, 34(8): 1121-1127.)
[4] Maul W A, Melcher K J, Chicatelli A K, et al. Sensor Data Qualification for Autonomous Operation of Space Systems[R]. AIAA FS-06-07.
[5] Melcher K J, Fulton C E, Maul W A, et al. Development and Application of a Portable Health Algorithms Test System[R]. NASA/TM 2007-214840.
[6] Hines J, Uhrig R. Use of Autoassociative Neural Networks for Signal Validation[J]. Journal of Intelligent & Robotic Systems, 1998, 21(2): 143-154.
[7] 崔文斌, 叶志锋, 彭利方. 基于信息融合遗传算法的航空发动机气路故障诊断[J]. 航空动力学报, 2015, 30(5): 1275-1280.
[8] Wong E, Fulton C E, Maul W A, et al. Sensor Data Qualification System (SDQS) Implementation Study[R]. NASA/TM 2009-215442.
[9] Guo-Jian H, Gui-Xiong L, Geng-Xin C, et al. Self-Recovery Method Based on Auto-Associative Neural Network for Intelligent Sensors[C]. Jinan: 8th World Congress on Intelligent Control and Automation(WCICA), 2010.
[10] Reyes J, Vellasco M, Tanscheit R. Measurement Correction for Multiple Sensors Using Modified Autoassociative Neural Networks[J]. Neural Computing & Applications, 2013, 24(7-8).
[11] 刘浩然. 一种基于改进遗传算法的神经网络优化算法研究[J]. 仪器仪表学报, 2016, 37(7): 1573-1580.
[12] 张书刚, 郭迎清, 陆军. 基于GasTurb/MATLAB的航空发动机部件级模型研究[J]. 航空动力学报, 2012, 27(12): 2850-2856.
[13] Guo T, Mattern D, Jaw L, et al. Model-Based Sensor Validation for a Turbofan Engine Using Autoassociative Neural Networks[J]. International Journal of Smart Engineering System Design, 2003, 5(1): 21-32.
[14] Meskin N. Multiple-Model Sensor and Components Fault Diagnosis in Gas Turbine Engines Using Autoassociative Neural Networks[J]. Neural Computing & Applications, 2013, 24(7-8).
[15] 王清, 招启军. 基于遗传算法的旋翼翼型综合气动优化设计[J]. 航空动力学报, 2016, 31(6): 1486-1495.
[16] Jeffrey T Csank, Donald L Simon. Sensor Data Qualification Technique Applied to Gas Turbine Engines[R]. NASA/TM 2013-216609.
[17] Simon D L, Bird J, Davison C, et al. Benchmarking Gas Path Diagnostic Methods: A Public Approach[R]. ASME 2008- GT- 51360.
[18] Simon D L. Propulsion Diagnostic Method Evaluation Strategy (ProDIMES) User S Guide[R]. NASA/TM 2010-21.　　　　　　　　　　　　　*　收稿日期:2017-04-07；修订日期:2017-05-26。作者简介:吕升,男,硕士生,研究领域为航空发动机智能控制与健康管理技术。E-mail: 596562202@qq.com(编辑:朱立影)