A top-weighted classification method with expert ability characterization for failure mode and effect analysis

Abstract

Failure mode and effect analysis (FMEA) is a useful tool to assess the potential risks of a system and provide necessary correction suggestions. This study proposes a top-weighted classification method with expert ability characterization for FMEA to deal with two crucial issues: few studies have attempted to characterize the ability of FMEA experts to provide accurate assessments in the risk assessment process, and the realistic feature that different risk categories have different importance was not considered. First, we develop a synergy theory-based weight iterative algorithm, by which the ability of experts is characterized and specific weight information is automatically generated at the element level of their assessment matrices. Then, we suggest a novel top-weighted distance measure that considers the importance of different risk categories, based on which the consensus-based top-weighted classification method (CTWCM) is proposed. After that, a simulation comparison experiment is designed to examine the performance of the CTWCM against the ABC analysis and ELECTRE-TRI methods. The numerical results show that the CTWCM outperforms the other two methods on all three key classification indices. In addition, a theoretical comparison is further presented to demonstrate our novelty and significance. Finally, the proposed method is illustrated through a SARS-CoV-2 management case.