BDWRPN: Belief divergence weighted risk priority number for failure modes ranking and its application

Ranking failure modes in complex system is important for analysis, diagnosis and prevention of risks in practical engineering. Failure mode and effects analysis (FMEA) theory provides the risk priority number (RPN) for ranking of failure modes. However, researches find that some shortcomings exist in the classical RPN methods of FMEA. For example, the uncertainty among FMEA experts is not modeled properly. To address this issue, we propose the belief divergence weighted risk priority number (BDWRPN) for failure modes ranking in FMEA in the framework of Dempster-Shafer evidence theory (DST). The belief divergence measure in DST is adopted to measure the uncertainty judgements from FMEA experts. The Deng entropy in DST is designed to measure the relative importance among FMEA experts. The Dempster's combination rule in DST is adopted for fusion of assessments on each risk factor coming from different FMEA experts. A case study of failure modes ranking for a sheet steel production process is designed to verify the effectiveness of the proposed method.