Fuzzy Bayesian Network Fault Diagnosis Method Based on Fault Tree for Coal Mine Drainage System

With the increase of the structure and function of the coal mine drainage system, there is often a strong coupling relationship between the faults of each component, which brings some difficulties to the fault diagnosis of the system. The existing fault diagnosis methods of coal mine drainage system do not consider the integrity of the system, and are only suitable for fault diagnosis of component parts. Combining the advantages of fault tree (FT) and Bayesian network, the fault diagnosis method of Bayesian network based on fault tree is proposed and its description ability, interpret ability and reasoning are improved. Aiming at the uncertainty of correlation strength between faults and symptoms of coal mine drainage system, fuzzy set theory is introduced into Bayesian network. The conditional probabilities between nodes are determined by the fuzzy algorithm. The system-level fault diagnosis model based on fuzzy Bayesian network (FBN) is constructed according to the mapping relationship between the fault tree and the Bayesian network. The causal inference and diagnosis analysis of mine drainage system are carried out by modeling and inference software. Taking the fault diagnosis of the coal mine drainage system in Shaanxi province as engineering application example, the coal mine drainage monitoring and fault diagnosis system is developed based on DSP and PC, and the feasibility and accuracy of the fault diagnosis model are verified. The results show that the average diagnosis accuracy of the fuzzy Bayesian network model is 83.5%, which is 10.3% higher than the traditional Bayesian network model.