Abstract This paper proposes a description rule and an evolutionary algorithm of a dumbbell-shaped multi-echelon supply chain complex network based on node degree preference and the connection preferences of competitive advantage coefficient. Furthermore, an evolutionary model of synthesis network and its subnets of supply chain is established to reveal the evolution of statistical characteristics. The model can reveal the correlation between the synthesis network and its subnets, and is verified by an example. A global network-based and core-echelon failure strategy is proposed based on the model. A node cascading failure algorithm is designed to measure dynamic robustness of a network, and a method for comprehensively evaluating robustness of supply chain complex networks from multiple dimensions is proposed. The impact of dynamic robustness on a supply chain is demonstrated. The research provides a scientific basis for the optimization of multi-echelon supply chain network structure and dynamic robust control.

中文翻译：

---

多级供应链复杂网络演化模型构建及级联失效鲁棒性分析

摘要 提出了一种基于节点度偏好和竞争优势系数连接偏好的哑铃形多级供应链复杂网络的描述规则和进化算法。此外，还建立了合成网络及其供应链子网的演化模型，以揭示统计特征的演化。该模型可以揭示综合网络与其子网之间的相关性，并通过实例进行验证。基于该模型提出了一种基于全局网络的核心梯队故障策略。设计了一种节点级联失效算法来衡量网络的动态鲁棒性，提出了一种多维度综合评价供应链复杂网络鲁棒性的方法。展示了动态稳健性对供应链的影响。该研究为多梯队供应链网络结构优化和动态鲁棒控制提供了科学依据。