基于节点关联度的生命线系统级联失效建模

发布时间：2018-06-24 20:42

本文选题：生命线工程 + 级联失效　；参考：《大连理工大学》2013年硕士论文

【摘要】：生命线系统是维系现代城市基本功能和区域经济的基础性工程设施,是城市的动脉和神经。生命线工程关系着国家安全和利益,对国家及地区正常管理功能的发挥具有重大意义。然而生命线工程多为内部关联紧密的网络系统,灾害作用下的生命线系统的破坏模式多为某处的一个微小扰动,通过节点之间的连接关系引起其他点发生故障,触发造成整个系统的连锁反应,最终导致系统中的大部分甚至整个网络的崩溃的灾难性后果。从静态角度研究系统的失效问题是远远不够的,有必要从动态的角度,建立一个关于生命线系统级联失效的普适性模型,研究其在灾害情形下的运行机制,从而为生命线工程的保护,政府部门的防灾减灾提供确定的方向和有理有据的决策支持。本文在综述国内外生命线工程研究历史和现状的基础上,引入节点综合重要度概念,从节点在网络结构中的空间位置和功能作用两个方面,利用可变模糊聚类迭代模型对生命线网络节点单元的重要度进行分析,确定网络关键节点。而后依据组织结构中的相关概念,把生命线系统的节点关系形式划分为直接相连的两个节点间下游节点对上游节点的直接从属关系、直接相连的两个节点间上游节点对下游节点的直接控制关系、间接相连的两个节点间下游节点对上游节点的间接从属关系、间接相连的两个节点间上游节点对下游点的间接控制关系,处在并列支路上的高级别上并列、同级别上并列关系几种形式,利用相邻模糊标度法定量计算出各种节点连接形式对应的关联因子,综合网络整体关联性,确定节点关联度,建立基于多维关联的生命线系统节点关联度模型。在灾害蔓延动力学理论基础之上,结合生命线系统的网络结构特征,利用节点关联度对已有蔓延动力学模型中的节点连接强度系数进行改进。从系统单元间的关联关系和考虑时间因素的动态传播入手,全面考虑生命线系统这类实体网络中资源分配的特征,建立一个生命线网络级联失效模型,研究灾害在网络的动态传播问题,考察节点连接形式和网络结构特征对生命线网络级联失效规模的影响。最后,通过对一段供水网络的分析和数值模拟研究,演示了本论文所构建模型的使用过程。实验结果表明,节点间的关联度放大了系统级联失效,而且当关键节点间接方式多为直接从属与控制等高关联度的网络发生故障时,网络级联失效效果更加明显,说明节点关联度是影响生命线系统灾害蔓延情况的重要因素。并且模拟结果与这些实际生命线系统的特征一致,表明所建立的模型可以有效地模拟生命线系统级联失效过程,给未来的生命线工程保护工作提供理论依据。
[Abstract]:Lifeline system is the basic engineering facility to maintain the basic function and regional economy of the modern city, it is the artery and nerve of the city. The lifeline engineering is related to the national security and interest, and is of great significance to the function of the normal management of the country and the region. The failure mode of the lifeline system is a small disturbance in a certain place, which causes the failure of other points through the connection between nodes, triggering the chain reaction of the whole system, and eventually causing the catastrophic consequences of the collapse of most and even the entire network. Far from enough, it is necessary to establish a universal model for the cascading failure of lifeline systems from a dynamic point of view, and to study the operating mechanism of the lifeline system in a disaster situation, thus providing a definite direction and a reasonable decision support for the protection of lifeline engineering and the disaster prevention and reduction of the government departments.
On the basis of the history and present situation of the research of lifeline engineering at home and abroad, this paper introduces the concept of node comprehensive importance. From the two aspects of the spatial position and function of the node in the network structure, this paper uses the variable fuzzy clustering iterative model to analyze the weight of the lifeline network node unit, and then determines the key node of the network. According to the related concepts in the organization structure, the node relationship form of the lifeline system is divided into direct subordinate relations between the two nodes directly connected to the upstream nodes, the direct control relationship between the upstream nodes of the two nodes directly connected to the downstream nodes, and the upstream nodes connected by the downstream nodes of the two nodes indirectly connected to the upstream nodes. The indirect subordinate relation of point, the indirect control relationship between the upstream nodes of the two nodes which are indirectly connected, the high level on the parallel branch and the parallel relationship in several forms. The node association degree is determined, and the node association degree model of lifeline system based on multidimensional association is established. Based on the theory of disaster spread dynamics, combining the network structure characteristics of lifeline system, the node connection intensity coefficient in the existing spread dynamics model is improved by the node association degree. The relationship between the system units is obtained. Considering the dynamic propagation of time factors, we consider the characteristics of resource allocation in lifeline systems, establish a lifeline network cascading failure model, study the dynamic propagation of disaster in the network, and investigate the effect of node connection and network structure characteristics on the cascading failure scale of lifeline network.
Finally, through the analysis and numerical simulation of a section of water supply network, the application process of the model is demonstrated. The experimental results show that the correlation degree between nodes amplifies the cascading failure of the system, and the network cascade is lost when the key nodes are mostly directly subordinate and controlled by the high correlation network. The effect is more obvious, indicating that the node association degree is an important factor affecting the lifeline system disaster spread, and the simulation results are consistent with the characteristics of these actual lifeline systems. It shows that the model can effectively simulate the cascading failure process of lifeline system, and provide a theoretical basis for the future lifeline engineering protection work. According to it.
【学位授予单位】：大连理工大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TU99;X913.4

【参考文献】