应急物流关联网络级联失效机理与关键节点抗毁性研究

A new situation appears generally in recent cases of disasters response. In emergency logistics, global collapse resulting from single node failure causes rescue delayed and even expands losses. Therefore, in this project we construct an interdependent networks model to protect emergency logistical system from cascading failure. By mapping functions of the inter layer networks, those associated networks including physical layer, informational layer and organizational layer are coupled in weights. Then, we analyze the transfer mechanism of node failure among the heterogeneous networks. Accordingly, we use both central and destructive indicators to build node importance function to identify the key nodes of interdependent networks, and put forward system optimization strategies against vulnerability in the end. Different from existing researches, during the process of analyzing cascading failure mechanism, we take node capacity restriction as well as physical loads of inner layer networks and structure loads of inter networks into consideration. Meanwhile, according to the timeliness requirement of emergency relief supplies distribution, the improved reallocation rule of failure node load is illustrated. With regard to the random association between nodes, we also make definition and supplement to the node failure condition of interdependent networks. The study of this project will be helpful in inspiring the emergency decision maker to think about the risks of emergency logistics operation systematically, and also have important theoretical value and practical significance to increase the operation reliability of emergency logistics system.

在以往突发灾害事件中，出现了因个别节点失效引发应急物流系统全局崩溃进而延误救援、造成损失扩大的现象。为了避免上述级联失效行为，本项目首先通过构建层间网络映射函数，建立应急物流关联网络模型，实现物理层、信息层与组织层网络的加权耦合。在此基础上，对节点失效在关联异质网络间的传递机理进行解析，并据此采用中心性与破坏性指标构建节点的重要性测度函数，对关联网络中的关键节点进行识别，从而提出应急物流系统优化的抗毁性策略。区别于现有研究，本项目在分析关联网络级联失效机理时，同时考虑了节点容量以及节点在同层网络中的物理负载和在网络间的结构负载；并基于应急救援物资配送的时效性要求，改进失效节点的负载重分配规则；针对节点间的随机关联特征，对关联网络节点的失效条件进行补充和界定。项目的研究成果将有助于启发应急决策者对应急物流运作风险进行系统性思考，对提高应急物流系统运作的可靠性具有重要的理论价值与实际意义。

在以往突发灾害事件中，出现了因个别节点失效引发应急物流系统全局崩溃进而延误救援、造成损失扩大的现象。为了避免上述级联失效行为，本项目首先通过构建层间网络映射函数，建立应急物流关联网络模型，实现物理层、信息层与组织层网络的加权耦合。在此基础上，对节点失效在关联异质网络间的传递机理进行解析，并据此采用中心性与破坏性指标构建节点的重要性测度函数，对关联网络中的关键节点进行识别，从而提出应急物流系统优化的抗毁性策略。区别于现有研究，本项目在分析关联网络级联失效机理时，同时考虑了节点容量以及节点在同层网络中的物理负载和在网络间的结构负载；并基于应急救援物资配送的时效性要求，改进失效节点的负载重分配规则；针对节点间的随机关联特征，对关联网络节点的失效条件进行补充和界定。项目的研究成果将有助于启发应急决策者对应急物流运作风险进行系统性思考，对提高应急物流系统运作的可靠性具有重要的理论价值与实际意义。