Reliability Analysis Of Multi-State Hydraulic System Based On T-S Importance And Bayesian Network

The hydraulic systems are widely used in various types of equipment and have an important core role in control and power transmission. The reliabilities of hydraulic systems are always been the key of industry technology to improve urgently. Therefore, the researches on the analysis methods of hydraulic system reliability have an important theoretical and practical significance. Because the existing hydraulic system reliability analysis methods are still limited to the conventional binary analysis methods, the establishing models of reliability analysis are quite different from the real conditions. To this end, this paper is dedicated to the research on the reliability analysis methods of the polymorphism hydraulic system and proposes the algorithms of T-S importance and of T-S fuzzy fault tree with the combination of Bayesian networks to promote the development of reliability methods and their applications in hydraulic engineering.Aiming at the different known conditions of system, three analysis algorithms of T-S importance are presented on the basis of the methods of T-S fuzzy fault tree analysis, including T-S state importance, T-S structure importance and T-S fuzzy importance. It is proved that the proposed methods of T-S importance are available by comparing the existing importance algorithms of traditional fault tree and fuzzy fault tree and combining with an example of hydraulic press's oil supply system. These three importance methods inflect the contribution of components to system in different views and provide evidence to reliability engineering applications in different conditions.The method of T-S fuzzy fault tree disposed by Bayesian network is presented on the research of binary and multi-state analysis methods of system reliability based on T-S fuzzy fault tree, Bayesian network and Barlow method. Aiming at the establishing of Bayesian network, the transformation approach of T-S fuzzy fault tree to the Bayesian network is presented. Then the algorithms of comparing Bayesian network with T-S fuzzy fault tree and Bayesian condition probability and T-S importance are finished. It is proved that the proposed method is available.A certain hoisting system of girder hoister is taken as an example for the analysis on importance algorithm of T-S fuzzy fault tree and Bayesian network and the reliability of hoisting system is improved.