航空发动机故障诊断的机载自适应模型

On board self-tuning model for aero-engine fault diagnostics

College of Energy and Power Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, China

摘要: 提出了复合拟合法建立状态变量模型,该方法应用于建立高维状态变量模型时,具有较高的精度.将健康参数作为增广的状态变量,设计了卡尔曼滤波器,从而可以根据可测参数的偏离量估计得到健康参数.为了减少自适应模型与真实发动机之间的建模误差,在自适应模型中加入神经网络对稳态基点模型进行修正,从而提高了故障诊断系统的置信度.
- 航空、航天推进系统 /
- 航空发动机故障诊断 /
- 健康参数 /
- 机载自适应模型 /
- 状态变量模型 /
- 卡尔曼滤波器 /
- 神经网络
Abstract: The composite least-square method was proposed for establishing the state variable model.The method was also applied to develop high dimension state variable model with high accuracy.By taking health parameters as augmented state variables,a Kalman filter was then designed to predict the health parameters from the deviation of measurable parameters.To minimize the modeling errors between the self tuning model and real engine,a neural network was built to modify the steady-state modeling errors,thus improving the confidence level of fault diagnosis system.
- aerospace propulsion system /
- aero-engine fault diagnostics /
- health parameters /
- on board self tuning model /
- state variable model /
- Kalman filter /
- neural network

[1]	Volponi A J,DePold H,Ganguli R.The use of Kalman filter and neural network methodologies in gas turbine performance diagnostics:a comparative study[J].Journal of Engineering for Gas Turbines and Power,2003,125(11):917-924.
[2]	Luppold R H,Gallops G W,Kerr L J,et al.Estimating in-flight engine performance variations using kalman filter coneepts[R].AIAA-89-2584,1989.
[3]	王信德,孙健国,李松林,等.神经网络在发动机自适应建模中的应用研究[J].航空动力学报,2003,18(6):845-849.WANG Xinde,SUN Jianguo,LI Songlin,et al.Application of neural networks to aeroengine adaptive modeling[J].Journal of Aerospace Power,2003,18(6):845-849.
[4]	冯正平,孙健国,黄金泉,等.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,1998,18(4);435-438.FENG Zhengping,SUN Jianguo,Huang Jinquan,et al.A new method for establishing a state variable model of aeroengine[J].Journal of Aerospace Power,1998,13 (4):435-438.
[5]	冯正平.航空发动机多变量鲁棒控制[D].南京:南京航空航天大学,1999.