ABSTRACT Any shortage and deviation in the implementation of logistics processes can lead to a poor performance and losing the market share of the manufacturer. To overcome this problem, the identification of existing risks in all logistics sub-processes, and taking corrective/preventive measures, is of great importance. In this study, a risk prioritization approach is presented based on a sequential multi-stage fuzzy cognitive map (SMFCM) and process failure mode and effects analysis (PFMEA) to prioritize logistics processes risks. This approach, in addition to considering a process-oriented view, prioritizes the logistics risks according to the amount of each risk’s impact on other risks through the internal and external-stage causal relationships as well as the values of risk factors. The score obtained from this approach can create enough distinction among the priorities of different risks in comparison with conventional risk priority number (RPN). Also, the results based on this score have a lower dependency on experts’ opinions by applying the Extended Delta Rule (EDR) learning algorithm. In fact, the intelligent nature of this approach enables decision-makers to identify critical risks and examine the status of the system at any time. The results of the implementation of the SMFCM-PFMEA approach in a manufacturing company active in the automotive spare parts industry confirm the superiority of this compared with the conventional score.

中文翻译：

---

基于连续多阶段模糊认知图的制造业物流过程风险因果效应分析：案例研究

摘要 物流流程实施中的任何短缺和偏差都会导致绩效不佳并失去制造商的市场份额。为了克服这个问题，识别所有物流子流程中存在的风险并采取纠正/预防措施非常重要。在这项研究中，提出了一种基于顺序多阶段模糊认知图 (SMFCM) 和流程故障模式和影响分析 (PFMEA) 的风险优先排序方法，以对物流流程风险进行优先排序。这种方法除了考虑面向过程的观点外，还通过内部和外部阶段的因果关系以及风险因素的值，根据每种风险对其他风险的影响程度来对物流风险进行优先级排序。与传统的风险优先级数 (RPN) 相比，从这种方法获得的分数可以在不同风险的优先级之间产生足够的区别。此外，通过应用扩展增量规则 (EDR) 学习算法，基于此分数的结果对专家意见的依赖性较低。事实上，这种方法的智能特性使决策者能够随时识别关键风险并检查系统状态。在一家活跃于汽车零部件行业的制造公司中实施 SMFCM-PFMEA 方法的结果证实了与传统评分相比的优越性。通过应用扩展增量规则 (EDR) 学习算法，基于此分数的结果对专家意见的依赖性较低。事实上，这种方法的智能特性使决策者能够随时识别关键风险并检查系统状态。在一家活跃于汽车零部件行业的制造公司中实施 SMFCM-PFMEA 方法的结果证实了与传统评分相比的优越性。通过应用扩展增量规则 (EDR) 学习算法，基于此分数的结果对专家意见的依赖性较低。事实上，这种方法的智能特性使决策者能够随时识别关键风险并检查系统状态。在一家活跃于汽车零部件行业的制造公司中实施 SMFCM-PFMEA 方法的结果证实了与传统评分相比的优越性。