A new method to evaluate risk in failure mode and effects analysis under fuzzy information

Failure mode and effects analysis (FMEA) is a useful and effective tool to identify and mitigate project risk, which utilizes the risk priority number (RPN) to determine the risk priority order of failure modes. In many applications when multiple experts give their opinions about one failure mode, the risk evaluations can be vague and imprecise, which could arise conflicting evidence that is hard to manage. To address this issue, the information offered by experts should be analyzed under a model of fuzzy numbers and Dempster-Shafer (D-S) combination theory. Here, the traditional RPN is not sufficient for risk evaluation. A new RPN is proposed in this paper with two parts. The first part is a product of memberships whose average degrees are equal to one, and the second part results from applying the Dempster-Shafer theory with tools of evidential downscaling method and belief entropy function. The new RPN can be effective and convictive to handle conflicting evidence in FEMA.