System Reliability Analysis Based On Copula And Determination Of Equipment’s Maintenance Period

In this paper, the previous Copula model is analyzed and its drawback is found. A composite Copula model is proposed based on the Copula theory to reflect different relevant information among random variables in mechanical system. The proposed model is applied to Bayesian Network to analyze the correlativity of bottom events in BN, which is helpful for assessing the reliability of mechanical system.(1) The reliability models of series system and parallel system is derived based on Copula theory. There is only some difference in the derived model instead of.multi dimensional integral. Copulas connect of commonly used model parameter estimation method is improved, and the improved method can estimate parametric with the help of Bayesian approach. A composite Copula model is proposed by a example. The composite model is more accurate than conventional one.(2) The structure of composite Copula can be confirmed based on maximum cardinality search algorithm. The composite Copula model is used in BN to consider the correlativity of bottom events, and the corresponding reliability computing method is proposed. Through the examples of composite copulas connect model can better describe the correlation between variables. Rank correlation coefficient of the linear correlation coefficient has the obvious advantage.(3) Based on cost and equipment reliability model of preventive maintenance cycle is proposed in which equipment preventive maintenance cost, maintenance cost of normal use, and reliability are considered To maximize earnings per unit time. This model is expanded, and is used to determine the equipment preventive maintenance cycle and replacement cycle. The expanded model is the foundation for building maintenance strategies model more complex system including equipments.