Fault diagnosis about aviation-engine based on information integrating

Abstract

The article discussed the advantage and disadvantage about neural network and expert system. According to the feature of fault for aviation engine, the paper puts forward the system of multi-parallel neural networks which are harmonized by expert system based on information integrating for the fault diagnosis of aviation engine. The expert system is used to monitor the work condition of aviation engine and harmonize the operation of neural networks. The neural networks are used to diagnose fault according to the order of expert coordination system. The system uses the modularity structure and is programmed by C and MATLAB language, the neural network toolbox is used to develop BP neural network which diagnose the fault. The system takes advantage of neural network and expert system rand overcomes their disadvantages. The simulating test and practical use show that the accuracy and work efficiency of fault diagnosis are greatly increased.