Abstract:Due to the complexity of the fault tree, the fault tree will firstly be decomposed to obtain sub-modules with similar modular structure, and also the module library will be built to store the sub-modules. This paper will introduce the basic processes and principles of feature-retrieval based on module library. In order to avoid repetition retrieval and NP problem, the concept of valid module will be introduced, that is to not only consider the hierarchy relationship of the event in fault tree, but also consider the affiliation between each event when traversing module library. Finally, this paper studies an optimal sorting method of importance based on Bayesian theory and also makes the importance correction. The core component of this paper is to build the module library based on the basic modular structure, and also introduce the concept of valid module as well as optimal sorting of importance based on Bayesian theory on the minimum cut sets. The present study result indicates that this method can effectively reduce the solving size, accelerate the speed of diagnosis, and improve the credibility of importance sorting as well.