Research On Assessement Of Node Enterprises Risk Propagation Capability Of Supply Chain Network

With the deepening of world economic integration and informatization and the deep integration of global production,the structure of the supply chain has undergone tremendous changes,having evolved from a simple linear single-chain structure to a complex dynamic network structure.It is accompanied by the frequent occurrences and rapid spread of risks throughout the supply chain network,which affects and even disrupts the normal operation of the entire supply chain.Moreover,the ability of node enterprises to influence other enterprises and supply chains varies greatly when risks occur.Therefore,it is the purpose and intention of this paper to provide the basic theoretical support for enterprises to carry out risk prediction and maintain the stable operation of the entire supply chain through identifying the best way to carry out an effective assessment of enterprises' risk level,discovering the "super node" in risk propagation,and scientifically assessing its ability to spread risks.This dissertation breaks away from the microscopic perspective of the past that pays too much attention to the factors of the enterprise itself.Under the framework of complex network theory,and specifically from the perspectives of network structure and propagation dynamics,issue of risk spreading capability assessment of node enterprise in the supply chain network is studied by abstracting enterprises into nodes,using the supply and demand relationship between enterprises as the supply chain network,and comprehensively applying theoretical analysis,modeling,simulation,case analysis and other methods The main work of the research is as follows:(1)This dissertation analyzes In-depth the formation mechanism of the supply chain network and the propagation process and dynamic characteristics of the supply chain risk.Based on the analysis of the mechanism of supply chain network formation,the theory of infectious disease model in propagation dynamics is selected as the theoretical basis of supply chain risk propagation,and the specific SIRS model is taken as an example to carry out theoretical extrapolation and simulation analysis for the risk propagation process of supply chain network.Furthermore,the characteristics of risk propagation capability are simulated from the perspective of node infectivity and interaction between nodes.After analysis and calculation,a clear correlation is found between the propagation ability of the node enterprises and the network structure characteristics in the process of risk propagation.(2)There are many risk factors affecting the supply chain node enterprises.By analyzing the risk factors affecting the supply chain,risks are divided into two categories: internal and external risk factors,and key risk factors are selected to establish a node enterprise risk indicator system.In order to determine the risk weight of each factor and then judge the risk level,the combination of analytic hierarchy process and fuzzy comprehensive evaluation is used to combine the qualitative and quantitative analysis to give the membership degree evaluation sets of different supply chain nodes belonging to a certain risk level.The risk level of the node is obtained,and the VR(Virtual Reality)video live broadcast industry supply chain network is taken as an example to verify the example.(3)Based on the structural characteristics of the supply chain network,this dissertation analyzes the impact ability indicators of the node enterprises in the supply chain risk propagation,and proposes representative local and global network structure features with a total of eight factors.Due to the certain correlation between features,the multi-dimensional feature index is reduced and optimized by PCA(Principal Component Analysis),and BP(Back Propagation)neural network is used to construct the evaluation model of node risk propagation capability level.Through the combination of historical data and expert scoring,the example verification is carried out.The results show that the evaluation model is reasonable and effective and has high precision.(4)Given the dynamic change of the supply chain network structure when risks occur,and considering the factors of the time series,risk type and attribute characteristics of the node enterprises,the tensor decomposition is used as the high-order subspace analysis method to conduct a risk propagation ability rating.Compared with traditional methods of vector or matrix analysis,the method of tensor modeling of the three-dimensional attribute features of nodes overcomes the shortcomings of traditional two-dimensional data analysis or forced reduction of high-dimensional data to two-dimensional.The nonlinear relationship in the enterprise data structure of the network node in dynamic supply chain is not damaged in the tensor decomposition process,which can more accurately evaluate the propagation capability level of the node enterprise in the dynamic network,and can also mine the role evolution of the node.Based on the node enterprise's own risk level assessment,this dissertation starts with the supply chain risk propagation mechanism,and evaluates the node enterprise risk propagation capability multi-dimensionally from the perspective of network topology,and further expands to the evaluation of the node enterprises risk propagation capability in the dynamic network by adopting a step-by-step approach and in-depth analysis.The research conclusions have important theoretical and practical significance for enterprises to define risk location,strengthen risk prevention and guide supply chain organizations to carry out emergency management to reduce economic losses caused by risk events.