Algorithms of Joint Failure Importance of Two Components

Weixing Bi & Jianjun Chen (2009) proposed a new algorithm of Joint Failure Importance (JFI), The new JFI is a useful metrics to rank components regarding their impact on system performance. However their definitions regarding JFI were limited to statistically independent component states. This paper removes the statistical independence restriction by showing that correct results hold in the more general case where component failures can be statistically dependent. in this paper: 1) gives explicit formulas for the improved JFI. These utility importance could identify and rank the most important components correctly within the system, 2) proves the improved JFI is always non-negative for some classes of systems, 3) illustrates computed results with numerical examples.