Abstract: Aiming at the existing method of fault diagnosis and its limitations,a model of expert(system) of fault diagnosis based on rough set and neural network was put forward in the paper,which was used to diagnose fault of double local-fans system in high-gas mine.The structure and the design of each functional model of the expert system of fault diagnosis of double local-fans in high-gas mine based on rough set and neural network were introduced stressly.The key technologies of the model of knowledge acquirement and repository of neural network were discussed,and the simulation result of the system was given in detail.