Liu Zheng, Mou Xun, Liu Hu-Chen, Zhang Ling
IEEE Trans Cybern. 2023 Mar;53(3):1566-1577. doi: 10.1109/TCYB.2021.3105742. Epub 2023 Feb 15.
Failure mode and effect analysis (FMEA) is a widely used reliability management technology to evaluate the risk of potential failures in a system, product, or service. Nevertheless, the normal risk priority number (RPN) method has been extensively criticized for many deficiencies in practical applications. To overcome the drawbacks of traditional FMEA, plenty of methods have been suggested in previous studies. But majority of them evaluated the risk factors of each failure mode directly and cannot take group and individual risk attitudes into account. In this article, we put forward a new FMEA approach integrating probabilistic linguistic preference relations (PLPRs) and gained and lost dominance score (GLDS) method. The PLPRs are adopted to describe the risk evaluations of experts by pairwise comparison of failure modes. An extended GLDS method is introduced to derive the risk ranking of failure modes considering both group and individual risk attitudes. Moreover, a two-step optimization model is proposed to determine the weights of risk factors when their weighing information is unknown. Finally, a load-haul-dumper machine risk analysis case is presented to demonstrate the proposed FMEA. It is shown that the approach being proposed in this study provides a practical and effective way for risk evaluation in FMEA.
失效模式与效应分析（FMEA）是一种广泛应用的可靠性管理技术，用于评估系统、产品或服务中潜在失效的风险。然而，常规风险优先数（RPN）方法在实际应用中因诸多缺陷而受到广泛批评。为克服传统FMEA的缺点，以往研究中提出了许多方法。但它们大多直接评估各失效模式的风险因素，无法考虑群体和个体的风险态度。在本文中，我们提出了一种将概率语言偏好关系（PLPRs）和得失优势得分（GLDS）方法相结合的新FMEA方法。采用PLPRs通过对失效模式的两两比较来描述专家的风险评估。引入一种扩展的GLDS方法，在考虑群体和个体风险态度的情况下得出失效模式的风险排序。此外，当风险因素的权重信息未知时，提出了一个两步优化模型来确定其权重。最后，给出了一个装载机风险分析案例以验证所提出的FMEA方法。结果表明，本研究提出的方法为FMEA中的风险评估提供了一种实用有效的途径。
