基于ReliefF-LMBP算法的涡轴发动机气路故障模式识别  

Gas Path Fault Mode Identification of Turboshaft Engine Based on Relief F-LMBP Algorithm

作　　者：王召广 杨宇飞 闫召洪 鲁峰[2] WANG Zhao-guang;YANG Yu-fei;YAN Zhao-hong;LU Feng(AECC Hunan Aviation Powerplant Research Institute,Zhuzhou 412002,China;College of Energy and Power Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China)

机构地区：[1]中国航发湖南动力机械研究所,湖南株洲412002 [2]南京航空航天大学能源与动力学院,江苏南京210016

出　　处：《推进技术》2021年第1期220-229,共10页Journal of Propulsion Technology

基　　金：国家科技重大专项(2017-I-0006-0007)。

摘　　要：针对涡轴发动机气路故障模式识别精度不高的问题,提出了一种基于ReliefF-LMBP故障特征提取的发动机故障模式识别方法。应用ReliefF算法对发动机传感器参数赋予权值,对传感器参数特征权重值进行迭代更新和排序,聚集好的特征样本,离散异类样本。根据筛选出的特征子集,利用LMBP神经网络算法进行发动机故障模式识别。以涡轴发动机为对象进行气路故障诊断验证,结果表明所提方法能提取特征传感器参数并实现有效的故障模式识别。In order to improve the performance of gas path fault diagnosis for turbo shaft engine,a Relief FLMBP based method was proposed to fault feature extraction and pattern recognition.Firstly,the available enginesensor measurements were analyzed and assigned feature weights,and the fault feature subsets were ordered anddetermined after iterative selection by Relief F algorithm.The effective feature measured parameters were gatheredby similar samples,and the rest parameters fell in the discrete heterogeneous sample subsets.Afterwards,theLMBP Neural Network algorithm was employed to build up the relationship between the fault modes and featuresof reduced measurements.The tests of gas path fault diagnosis are carried out on a turbo shaft engine,and resultsshow the capability of feature extraction and superiority of fault pattern recognition.

关 键 词：涡轴发动机 气路故障诊断 特征提取 神经网络 ReliefF分析 

分 类 号：V235.113[航空宇航科学与技术—航空宇航推进理论与工程]