PBGFMEA: A Belief Jensen–Shannon Theory and Gray Relational Projection-Based Enhanced FMEA Approach for Risk Evaluation of Integrated Circuit Process

Failure mode and effects analysis (FMEA) is a crucial risk assessment method widely used in the design of industrial products. However, a key challenge remains in handling the uncertainty and conflicts that arise from subjective evaluations. To address this, this paper introduces an enhanced FMEA model, termed PBGFMEA, which integrates Belief Jensen–Shannon theory and the Gray Relational Projection Method (GRPM). Belief Jensen–Shannon theory is utilized to quantify the discrepancies between Basic Probability Assignments (BPAs), effectively managing conflicting assessments in FMEA. To combine expert evaluations, Dempster’s combination rule is applied, generating a unified risk factor value. The resulting BPA fusion is then transformed into a probability distribution using an improved pignistic probability function. GRPM is employed to rank the risk priorities of failure modes, overcoming the issue in traditional FMEA where different failure modes might share the same risk priority. An application to integrated circuit manufacturing processes is provided, demonstrating the proposed method’s reliability and efficiency.