Probabilistic double hierarchy linguistic risk analysis based on failure mode and effects analysis and S-ARAS method

Abstract

Failure mode and effects analysis (FMEA) is regarded as useful and efficient tool in identifying and eliminating system failures and important means in safety and reliability analysis. It is widely utilized in diverse areas for improving and enhancing the performance of systems. However, in the procedure of risk assessment, traditional FMEA method usually ignores the influence of weights. Simultaneously, when represent risk assessment information, fuzzy and complex information cannot be represented accurately. Besides, psychological behaviors of decision-makers cannot be considered. Thus, this paper aims to present a new method based on FMEA and S-additive ratio assessment (S-ARAS) method to address the marine risk assessment problems under probabilistic double hierarchy linguistic (PDHL) environment. Firstly, the probabilistic double hierarchy linguistic term sets (PDHLTSs) are utilized to express the evaluation information of failure modes. The expert evaluation information of failure modes and subjective weight is also assessed by using PDHLTSs. The PDHLTSs can describe the information more reasonably and precisely. Secondly, FMEA method is used to evaluate these alternatives from three aspects of Severity (S), Occurrence (O) and Detection (D). Then, the bounded rationality of decision-makers (DMs) is considered. So, the prospect theory and ARAS method is combined to evaluate the risk problems. Meanwhile, comprehensive weight determination method is established based on the social network analysis (SNA) and The Integrated Determination of Objective CRIteria Weights (IDOCRIW) method under PDHL environment. The integrated method is more flexible in determining the importance of failure modes and avoids the randomness of subjective evaluation. Finally, the newly proposed method is applied to address the marine risk assessment problem. Sensitivity analysis of parameter and weight methods is carried out. Furthermore, comparison analysis is given to illustrate the superiority of the presented method.