Factorial Design Approach for Importance Analysis of System Reliability

The importance of system component parameters signifies their effect on system reliability. Most existing methods for importance analysis are computationally intensive, i.e., they require a large number of observations, or do not consider the features of complex dynamic systems. This study proposes a method to calculate the importance measure of component parameters on the basis of factors in factorial design. Two examples are presented, one for the structural importance of fault trees and the other for the importance of the component failure rate of a phased mission system (PMS). The proposed measure can take conventional structural importance as a special case and accurately reflect the characteristics of the PMS.