Parametric fault modeling and diagnostics of a turbofan engine

A parametric fault modeling and diagnostics approach of a turbofan engine is presented. The healthy engine model is obtained as a steady state map between input parameters representing engine operating conditions and output parameters governing engine performance characteristics. The fault modeling is conceptualized as either degraded engine performance levels or as sensor failures and both incipient and abrupt fault scenarios are considered. The fault parameters are successfully estimated using generalized least squares (GLS) techniques with the data obtained from several recorded flight data. Further, a nonlinear estimation method is introduced to improve the diagnostic performance in certain cases of large and abrupt failures.