Differential Importance Measure of Markov Models Using Perturbation Analysis

Reliability importance measures providing information about the importance of components on the system performance (reliability, maintainability, safety, or any performance metrics of interest) have been widely used in reliability studies and risk analysis. In this paper, the differential importance measure (DIM) introduced recently for use in risk-informed decision-making is extended to Markov reliability models. This allows to qualify the relative contribution of a component (or a group of components), as well as a state (or a group of states) on the total variation of system performance. The estimation of DIM at steady state from the operating feedback data by using perturbation analysis is also investigated. A numerical example of a dynamic system is finally introduced to illustrate the use of DIM, as well as the advantages of proposed approach.