An improved risk priority number model for FMEA based on belief measure

Decision making under uncertainty in risk analysis is a key issue in practical engineering. The risk priority number (RPN) model in failure mode and effects analysis (FMEA) is a widely used tool for ranking of risk items. However, there are limitations in the traditional RPN model. For example, it cannot represent the subjective or inaccurate judgements coming from FMEA experts. In addition, the uncertainty in the assessments of experts in FMEA item is not modelled and transformed to the RPN values. In this paper, a new risk priority number model based on belief Jensen-Shannon divergence measure and Deng entropy in Dempster-Shafer evidence theory is proposed. In the proposed method, the belief Jensen-Shannon divergence measure is adopted to effectively deal with the fuzziness and abnormal adjustment coming from all the FMEA experts. In addition, Deng entropy is used to quantify the uncertain degree of each expert and the result is modelled as a relative importance degree of expert. Dempster's combination rule is used to fuse experts' assessments of different failure modes to generate the integrated values of each risk factor. The rationality, superiority and effectiveness of the proposed RPN model are verified based on a case study of a production process in steel industry.