An ELICIT information-based ORESTE method for failure mode and effect analysis considering risk correlation with GRA-DEMATEL

Abstract

Failure mode and effect analysis (FMEA) is one of the most powerful reliability analysis techniques for identifying and preventing potential risks across various fields. Current FMEA methods, while effective, still present several shortcomings. For example, using experts’ subjective pairwise comparisons of risk factors to determine their weights reduces the stability of the result; different relationships among failure modes are often ignored. To improve the performance of FMEA, multi-criteria decision-making (MCDM) methods have been employed to support risk evaluation and prioritization in recent years. This paper proposes a novel FMEA method by exploring several MCDM techniques. First, the Extended Comparative Linguistic Expressions with Symbolic Translation (ELICIT) are utilized to generate group risk assessments under uncertainty. Then, grey relation analysis (GRA) is incorporated into the decision-making trial and evaluation laboratory (DEMATEL) method to objectively determine the weight of risk factors. Afterward, the traditional ORESTE (organísation, rangement et Synthèse de données relarionnelles (in French)) method is generalized to the ELICIT environment to prioritize the failure modes, in which the Besson’s ranking is replaced by deviation measure for more accurate ranking results. Finally, a case study of FMEA for electro-mechanical actuators is presented to illustrate the effectiveness of the proposed method. The results indicate that our approach can express risk information more flexibly, determine the weight of risk factors more objectively, and prioritize failure modes more reasonably.