Computation of Hybrid Uncertainty and Dependent Failure in System Reliability Analysis and Assessment

A reliability assessment framework is severely influenced by hybrid uncertainty, dependent failure and many other important factors. To this end, this paper proposed a more complete reliability assessment framework for computation of hybrid uncertainty and dependent failure. Bayesian Network is a common technique to model complex systems and propagates system information. On the basis of Bayesian Network, this paper use probability-box to represent hybrid uncertainty and Copula function to manipulate dependent failure. An availability case of a precision compensation loop of hydraulic system for heavy computerized numerical control machine tool is used to illustrate the applicability of our computation framework in practical problems.