Based On Fuzzy Neural Network Fault Diagnosis Of Expert Systems

This paper presented a compond intelligent equipment fault diagnosis system, and analyzed some failure analysis method, and determined the best scheme by experiment. From the results of simulation and experiments, the system has a good simulation performance. The research content as follows:First of all, after analyzed the research situation of the equipment fault diagnosis expert system at home and abroad and the shortcomings of the single expert system, the paper put forward the basic principle of the design. On this basis, it put forward a compond intelligent fault diagnosis system through the analysis of the processing manufacture enterprise's common equipment characteristics.Second, it made three kinds of Neural Network model by using the neural network technology. Then it compare and analyse the network model with MATLAB, and choose the most suitable neural network model after generating, solving specific design, training and the simulation the system.Again, it optimized the model parameters, and put the input and output expert knowledge in the form of fuzzy rules, and analyzed the rationality of diagnosis way by use the Fuzzy Neural Network theory.Finally, the man-machine interface can call the MATLAB function and Excel date by VB code.it establish fuzzy neural network model and apply to the enterprise field for the test. The results show that the system has many common characteristics and it's basic performance meet the design requirements.