Remote intelligent expert system for operation state of marine gas turbine engine

A distributed networked remote fault prognostic and diagnostic expert system for marine gas turbine is introduced which can realize cross-regional, multi-expert involved in collaborative decision-making mechanism. The expert system includes four layers namely the field data collection layer, the local condition monitoring layer, the network communication layer and the long-distance expert supports layer. The expert system uses artificial neural network to carry out real-time fault prognostic analysis for the operational status of key equipment to discover hidden or impending equipment faults, so as to effectively avoid the occurrence of the “lack of maintenance” and “excess maintenance”. The integration of fault diagnosis mechanism based on rough set and artificial neural network is used, which effectively solve the problems of typical fault diagnosis for a long time and a high false alarm rate. Finally, this paper describes the main characteristics and application of expert system to the remote fault prognosis and diagnosis of a gas turbine fuel system as an example for testing its capabilities and main features.