Abstract: In order to solve the disadvantages of difficulty to construct Bayesian network model and nodes’ conditional probability tables and to obtain the root nodes’ fault rates and fault probabilities accurately in Bayesian network method, and the shortage of complicated calculation and no reverse inference in T-S fault tree analysis method, a fuzzy reliability assessment method based on T-S fault tree and Bayesian network is proposed：the Bayesian network model and conditional probability tables are constructed by T-S fault tree and T-S gate rules respectively; the multi fault states of nodes are described by fuzzy numbers, the nodes multi fault states’ fault rates and fault probabilities are described by fuzzy subsets; the probability of leaf node’s fault states and root nodes’ state importance are proposed at the condition of just knowing root nodes’ fault states, and the fault rates and fault probabilities fuzzy subsets of leaf node’s fault states, and root nodes’ fuzzy importance and posterior probabilities at the condition of knowing the fault rates and fault probabilities fuzzy subsets of root nodes’ fault states are proposed by the Bayesian network inference. It is proved that the proposed methods are feasible by comparing with the T-S fault tree method in reference [5] and Bayesian network method in reference [10]. At last, fuzzy reliability assessment of coalmine roadway transporter’s hydraulic system is completed by the proposed method, the reliability indexes such as the root nodes’ state importance is calculated out, which provides basis for improving system reliability and completing fault diagnosis.

Key words: reliability assessment;Bayesian network;T-S fault tree;fuzzy subset;importance;hydraulic system