Expert Systems for Fault Diagnosis Integrating Neural Network and Fuzzy Inference

In order to improve the accuracy of diagnosis to satisfy the maintenance of weapon equipment, a kind of expert system (ES) is used in this paper. The system is integrated into neural network (NN) and fuzzy inference. The basic system structure diagrams adopt the frame of expert systems. ES is logic inference part and responsible for symbol processing. The status parameters are described by fuzzy theory and the structure model is built with fuzzy directed graph. The fuzzy logic inference provides an appropriate knowledge representation method to depict fuzzy knowledge. NN is responsible for numerical value calculation. The learning ability of NN can partially or entirely resolve the bottleneck problem of knowledge acquisition. The results indicate the validity and rationality of the model and the method.