System Reliability Analysis And Risk Priority Number Calculation Method Based On Copula Function

With the rapid development of science and technology,equipment and systems are becoming larger and more complex.In engineering practice,once the large-scale equipment and system operation failure,it will lead to varying degrees of economic losses,even will lead to heavy casualties.Therefore,the evaluation of reliability and safety of large equipment and complex system has important engineering application value.In general,the components of the system interact and influence each other,and operate in the same environment,so there is a strong nonlinear correlation between the components of the system.However,the existing system reliability analysis and risk assessment methods often assume that the components are independent of each other.Although this can bring convenience to the process of analysis and calculation,there is a large error in the analysis and calculation results,which affects the correctness of decision-making to a certain extent.Therefore,how to consider the correlation between system variables in reliability analysis is one of the difficulties to be solved.In view of the above existing problems,this paper uses the Copula function to model the correlation between variables and components,and studies the system reliability analysis and risk assessment method based on the Copula function.The main research contents are as follows:(1)An important degree calculation method for multi-state systems under related conditions.Importance calculation is used to find the weak link of the system in reliability engineering,which is of great significance to the reliability analysis of key components.In view of this,a new important degree calculation method under the condition of multistate system correlation is proposed,and the Copula function model is applied in the calculation process of the system comprehensive importance degree IIM.The relationship between component set C and system set S is connected by Bayesian formula,and the prior probability of components is used to calculate the posterior probability of the system and subsystems,so as to analyze and rank the importance of each component of the whole system.Copula function is used to model the correlation between components,which solves the problem that the reliability or structural characteristics of components are mainly concerned while the interaction and mutual influence of components are ignored in the process of importance calculation.(2)Reliability analysis of multi-state system component monitoring data fusion under correlation condition.Existing research on system reliability evaluation focuses on reliability analysis based on failure time,but this method cannot reflect the failure difference between different systems.To solve this problem,a reliability analysis method of multi-state system component monitoring data fusion under the condition of correlation was proposed,and the uncertainty of monitoring data was taken into account by Bayesian formula,so as to update the system reliability model with the real-time data obtained from monitoring.It not only improves the accuracy of individual reliability evaluation of the system,but also solves the problem of certain nonlinear correlation between components,which provides a strong premise for the maintenance and overhaul strategy of the system later.(3)Study on a new method for calculating the priority number of systemic risk considering weight under relevant conditions.USES the minimum cut set theory to simplify the large complex system,based on copulas connect function to modeling of components and the correlation of cut set,considering the parts and the incidence of minimum cut set of detection and severity degree of D S O,weight and hierarchical single sorting based on the analytic hierarchy process(ahp)and hierarchical total sorts,can effectively determine the components of the influence of system risk and failure mode.Because the proposed method is based on the minimum cut set and takes the correlation into consideration,the limitations of the traditional risk priority number method which assumes the independence between components and is only applicable to a single point of failure are solved.