Modelling methods for improving robustness in fault diagnosis of jet engine system

Modeling uncertainty is an inevitable consequence of the complexity of jet engine systems, and accurate dynamic models can never be fully obtained. The authors concentrate on the derivation of suitable mathematical models of a jet engine, to enable robust fault diagnosis designs to be achieved. The modeling uncertainty can be described as an additional term in the dynamic structure. Based on this structure, a robust fault diagnosis scheme can be designed. New methods are presented for determining the structure of modeling uncertainty. The work uses a 17th order nonlinear jet engine simulation model to illustrate the methods proposed. The simulation results show the power of the method in modeling jet engine systems for robust fault diagnosis purposes