基于自适应拟合建模的航空发动机健康参数估计
|
摘要
针对航空发动机各部件会发生性能缓慢退化,而现有的发动机线性模型建模方法无法在全寿命期限内满足精度要求,导致滤波算法对健康参数的估计误差随着飞行循环数的增加而增加的问题,提出自适应拟合的方法建立包含健康参数的状态变量模型(SVM),并设计卡尔曼滤波器进行健康参数估计。方法利用上一个采样周期内健康参数的估计值实时更新健康参数相关系数矩阵,以提高线性模型的精度。以某型商用航空发动机部件级模型为基础,在设计巡航点采用提出的方法建立线性变参数模型。通过相似换算,在多个飞行状态点,进行数字仿真性能缓慢退化过程。经验证:相比改进拟合法,基于所提建模方法的卡尔曼滤波器能较大地提高对健康参数的估计精度,同时具有较好的实时性。
Aiming at problem that components of aircraft engine will degrade slowly during its life cycle,but current linear model modeling methods of engine cannot meet requirement of precision for the whole life cycle,which leads to the fact that estimation errors of filtering algorithm for health parameters increases with flight cycles,an adaptive fitting method is proposed to establish state variable model( SVM) and Kalman filter is designed for health parameters estimation. The method uses health parameter estimation value to realtime update related coefficient matrices in the last sampling period,so as to improve precision of the linear model. Linear variable parameter model is established for a commercial aircraft engine,at design cruise point on the basis of component level model. And by applying similarity transformation,at several flight state points,performance degradations process are digitally simulated. The simulation results reveal that KF based on the proposed modeling method greatly improves the estimation precision of health parameters and also has good real-time ability.
Aiming at problem that components of aircraft engine will degrade slowly during its life cycle,but current linear model modeling methods of engine cannot meet requirement of precision for the whole life cycle,which leads to the fact that estimation errors of filtering algorithm for health parameters increases with flight cycles,an adaptive fitting method is proposed to establish state variable model( SVM) and Kalman filter is designed for health parameters estimation. The method uses health parameter estimation value to realtime update related coefficient matrices in the last sampling period,so as to improve precision of the linear model. Linear variable parameter model is established for a commercial aircraft engine,at design cruise point on the basis of component level model. And by applying similarity transformation,at several flight state points,performance degradations process are digitally simulated. The simulation results reveal that KF based on the proposed modeling method greatly improves the estimation precision of health parameters and also has good real-time ability.
引文
[1] Mercer C R,Simon D L,Hunter G W,et al. Fundamental technologydevelopment for gas-turbine engine health management[R].Washington:Glenn Research Center,NASA:2007.
[2] Litt J S,Simon D L,Gary S,et al. A survey of intelligent controland health management technologies for aircraft propulsionsystems[R]. Washington:Glenn Research Center,NASA:2005.
[3] Simon D L. An integrated architecture for on-board aircraft engineperformance trend monitoring and gas path fault diagnostics[R].Washington:Glenn Research Center,NASA:2010.
[4]马敏,周苗苗,李新建.基于ECT技术的航空发动机尾气监测系统设计[J].传感器与微系统,2015,34(5):88-91.
[5]马敏,闫超奇.基于LabVIEW的航空发动机气路故障分析系统设计[J].传感器与微系统,2017,36(8):103-105.
[6]赵汪洋,杨功流,庄良杰.多导航传感器的多模卡尔曼滤波研究[J].传感器与微系统,2006,25(7):85-88.
[7]郭淑霞,姜颖,刘佳.多传感器系统中基于扩展卡尔曼滤波器的在线故障检测[J].传感器与微系统,2015,34(6):126-129.
[8]江丹,邓新蒲,吴京,等.几何扩展卡尔曼滤波算法在卫星姿态确定系统中的应用[J].传感器与微系统,2016,35(9):156-160.
[9]姚依翔,谢俊元.基于UKF区域交叉定位的WSNs Sink节点动态跟踪算法[J].传感器与微系统,2015,34(4):123-126.
[10]王桂荣,李建勇.改进UKF算法在PMLSM无位置传感控制中的应用[J].传感器与微系统,2017,36(2):158-160.
[11]黄伟斌,张鹏,黄金泉.航空发动机状态变量模型建模方法研究[C]∥中国航空学会发动机自动控制学术交流会,贵阳:中国航空学会航空发动机自动控制专业委员会,2006:369-373.
[12]冯正平,孙健国.一种建立航空发动机状态变量模型的新方法[J].航空动力学报,1998,13(4):435-438.
[13]俞刚.涡扇发动机气路健康参数估计方法研究[D].南京:南京航空航天大学,2012:8-17.
[14]周文祥.航空发动机及控制系统建模与面向对象的仿真研究[D].南京:南京航空航天大学,2006:35-49.
[15] Litt J S,Parker K I,Chatterjee S,et al. Adaptive gas turbineengine control for deterioration compensation due to aging[R].Washington:NASA:Glenn Research Center,2003.
[16]黄伟斌,黄金泉.航空发动机故障诊断的机载自适应模型[J].航空动力学报,2008,23(3):580-584.
[17]殷锴,陶金伟,王鸿钧,等.民用航空发动机控制系统回路设计与仿真[J].航空计算技术,2012,42(6):107-110.
[18]佘云峰,黄金泉,鲁峰.基于卡尔曼滤波器的涡轴发动机气路部件性能估计[J].长春理工大学学报:自然科学版,2010,33(3):33-36.
[19]胡卫红,李述清,孙健国.控制问题中航空发动机飞行包线区域最优划分[J].推进技术,2011,32(3):391-395.
[2] Litt J S,Simon D L,Gary S,et al. A survey of intelligent controland health management technologies for aircraft propulsionsystems[R]. Washington:Glenn Research Center,NASA:2005.
[3] Simon D L. An integrated architecture for on-board aircraft engineperformance trend monitoring and gas path fault diagnostics[R].Washington:Glenn Research Center,NASA:2010.
[4]马敏,周苗苗,李新建.基于ECT技术的航空发动机尾气监测系统设计[J].传感器与微系统,2015,34(5):88-91.
[5]马敏,闫超奇.基于LabVIEW的航空发动机气路故障分析系统设计[J].传感器与微系统,2017,36(8):103-105.
[6]赵汪洋,杨功流,庄良杰.多导航传感器的多模卡尔曼滤波研究[J].传感器与微系统,2006,25(7):85-88.
[7]郭淑霞,姜颖,刘佳.多传感器系统中基于扩展卡尔曼滤波器的在线故障检测[J].传感器与微系统,2015,34(6):126-129.
[8]江丹,邓新蒲,吴京,等.几何扩展卡尔曼滤波算法在卫星姿态确定系统中的应用[J].传感器与微系统,2016,35(9):156-160.
[9]姚依翔,谢俊元.基于UKF区域交叉定位的WSNs Sink节点动态跟踪算法[J].传感器与微系统,2015,34(4):123-126.
[10]王桂荣,李建勇.改进UKF算法在PMLSM无位置传感控制中的应用[J].传感器与微系统,2017,36(2):158-160.
[11]黄伟斌,张鹏,黄金泉.航空发动机状态变量模型建模方法研究[C]∥中国航空学会发动机自动控制学术交流会,贵阳:中国航空学会航空发动机自动控制专业委员会,2006:369-373.
[12]冯正平,孙健国.一种建立航空发动机状态变量模型的新方法[J].航空动力学报,1998,13(4):435-438.
[13]俞刚.涡扇发动机气路健康参数估计方法研究[D].南京:南京航空航天大学,2012:8-17.
[14]周文祥.航空发动机及控制系统建模与面向对象的仿真研究[D].南京:南京航空航天大学,2006:35-49.
[15] Litt J S,Parker K I,Chatterjee S,et al. Adaptive gas turbineengine control for deterioration compensation due to aging[R].Washington:NASA:Glenn Research Center,2003.
[16]黄伟斌,黄金泉.航空发动机故障诊断的机载自适应模型[J].航空动力学报,2008,23(3):580-584.
[17]殷锴,陶金伟,王鸿钧,等.民用航空发动机控制系统回路设计与仿真[J].航空计算技术,2012,42(6):107-110.
[18]佘云峰,黄金泉,鲁峰.基于卡尔曼滤波器的涡轴发动机气路部件性能估计[J].长春理工大学学报:自然科学版,2010,33(3):33-36.
[19]胡卫红,李述清,孙健国.控制问题中航空发动机飞行包线区域最优划分[J].推进技术,2011,32(3):391-395.