A method based on q-rung orthopair fuzzy cognitive map and TOPSIS method for failure mode and effect analysis considering risk causal relationships

Abstract

As a preventive maintenance technique, failure mode and effect analysis (FMEA) is widely applied to determine and eradicate possible failures. In recent years, many FMEA methods have been developed to address the limitations of traditional FMEA methods in failure mode (FM) assessment, weight calculation, and FM prioritization in complex uncertain environments. However, due to the complexity of equipment, systems, services, etc., there is an inevitable influence on causal relationships between FMs, which is rarely considered by existing methods. Therefore, a novel FMEA method, which integrates the technique for order preference by similarity to the ideal solution (TOPSIS) of q-rung orthopair fuzzy set (q-ROFS) and q-rung orthopair fuzzy cognitive map (q-ROFCM), is proposed for reasoning the relationships between FMs. The design of the novel method, named q-ROFCM FMEA, involves three stages. The q-ROFS is applied to evaluate the risk factors for capturing more ambiguous information in the first stage. The q-ROFCM is proposed and the causal relationship between FMs is integrated to describe the impact of the relationships on the evaluation results through the dynamic behavior of FMs in the second stage. In the third stage, the TOPSIS method is applied to determine the priority ranking of FMs based on more reliable similarity measurement results of the projection measure. The proposed new method is applied to the FM risk assessment of power units in the railway turnout system. The competitiveness and effectiveness of the proposed method have been demonstrated by comparison with other classic methods and sensitivity analysis.