Aeroengine Gas Path Fault Diagnosis Using Rough Sets and Neural Networks

Aeroengine is a very complex nonlinear object.Traditional methods for its fault diagnosis were proved time-consuming and low efficient.A new system based on rough sets and neural networks for the fault diagnosis of aeroengine gas path component faults was presented in this paper.At first,the rough set theory was used to detect qualitatively faults and to isolate the fault.It consists of three steps:performing the discretion of sensed data,establishing the decision table and generating rules.After that,feed-forward neural networks are added into the system to construct several subsystems,which take the engine sensible data pretreated by rough sets as inputs and compute damage degrees of the aeroengine fault.Finally,the noise rejection abilities of the engine fault diagnosis system were analyzed.The test results show that the system can quantitatively diagnose the faults of aeroengine gas path components precisely and efficiently,while it is robust for noise rejection.