Research On Cascade Failure For Complex Logistics Network

With the rise and prosperity of the Internet era,e-commerce has gradually become the new normal,and the technological level of the logistics industry has been continuously promoted.As an emerging industry,the logistics industry has become the basic industry and arteries of the development of our national economy,and one of the important signs of comprehensive strength.Traditional logistics began in the era of the planned economy,and is mainly a storage and transportation company of various countries established by administrative divisions.In order to adapt to the development of technology and the ever-changing needs of customers,the traditional logistics industry has begun to transition to networked operations.However,the logistics network is extremely non-uniform,and this characteristic form an important feature of a complex logistics network-"robust but fragile."Because the operation of the logistics network is highly dependent on the external environment,emergencies,natural disasters,and“out of stock”shopping festivals pose a great threat to the normal operation of the logistics network.When an uncertain crisis occurs,the damaged network nodes are different,and the degree of cascading failure and the robustness of the network are also very different.When the ordinary logistics node fails,the logistics network still has high robustness;however,when the key logistics hub node fails,the robustness of the logistics network has great uncertainty.This kind of robust uncertainty may cause the logistics network to be paralyzed,cause great economic losses,and even determine whether the logistics network can continue to survive.Therefore,the traditional development model is gradually being subverted.In the 21st century,scientists have begun to actively explore high-efficiency,low-cost,and robust logistics systems.Compared with the traditional simulation research on the logistics network structure and cascading failure process based on theoretical assumptions,experts and scholars now pay more attention to modeling real logistics networks.However,the scale of the logistics network in reality is huge,and the data is intricate and complex,making field research even more impractical.Based on the theory of complex logistics networks,the characteristics of real logistics networks can be accurately revealed.The research results centered on controllability robustness are more practical and have far-reaching research value.In order to understand the logistics network in a deeper level and provide scientific methods and empirical support for the robustness of the network after the cascading failure,this article uses complex network theory to analyze and mine actual characteristics based on real logistics network data,and explore the complex characteristics of the logistics network.Network topology,load-capacity relationship,cascading failure process,and dynamic characteristics of load are used as the theoretical basis to build a more realistic complex logistics network model.For the first time,a complex logistics network cascading failure based on a nonlinear load-capacity model is proposed and a dynamic cascading failure model of complex logistics network is constructed.A series of cascading failure processes,cascading failure models,controllability robustness and economy,and dynamic cascading failure model comparison simulation analysis were performed to prove the advancement of the method and model of this paper.Optimality has great significance and far-reaching impact on solving practical logistics problems.The main research contents include the following five aspects:(1)Analysis,modeling and simulation verification of topological characteristics of logistics networks based on complex networksRelying on the real logistics network data of an enterprise,we analyze the network characteristic values and structural characteristics of complex logistics networks.The empirical counter degree distribution obeys the power law distribution and presents scale-free network characteristics.Therefore,combined with logistics aggregation and diffusion,we construct a complex logistics network model,and verify that it conforms to the structural characteristics of a real logistics network.This provids a theoretical reference for the further studying the complex logistics network model of cascading failures,and solving realistic logistics problems.(2)A cascading failure model for a complex logistics network based on a nonlinear load-capacity modelTaking the load-capacity data of a company as a research object,it is demonstrated that the load and capacity of the complex logistics network conform to a nonlinear relationship.Based on this,a nonlinear load-capacity model is constructed.By analyzing the parameter characteristics of the model,it is found that:Adjusting the parameter?,and?,we can obtain more accurate remaining capacity of logistics nodes,and the effect of parameter?on the relationship of the load-capacity is more significant than the effect of parameter?.Based on the nonlinear load-capacity model and the commonly used linear load-capacity model,four complex logistics network cascading failure models are constructed(Initial load redistribution strategy cascading failure model under linear load-capacity model,differential load redistribution strategy cascading failure model under linear load-capacity model,initial load redistribution strategy under nonlinear load-capacity model and nonlinear load-capacity model under differential load redistribution strategy model),the formulas for the load redistribution results after cascading failure are respectively derived.We can find that based on the linear load-capacity model,the initial load redistribution strategy and the differential load redistribution strategy after cascading failure load redistribution give the same consequence and based on nonlinear load-capacity model,the load redistribution results are different under the initial load redistribution strategy and the differential load redistribution strategy.The results show that the nonlinear load-capacity model is more controllable than the linear model against the cascading failure,providing more optimizable space for improving the robustness of complex logistics networks,and the nonlinear load-capacity model adopts the differential load redistribution strategy,which can more flexibly and rationally allocate the load of the failed nodes by adjusting parameter?,thereby enhancing the robustness of the complex logistics network.In summary,among the four cascading failure models of complex logistics networks,the differential load redistribution strategy model under the nonlinear load-capacity model is theoretically a better cascading failure model.(3)Analysis of cascading failure process based on real logistics networkAiming at the complex network abstracted by the real logistics network,the cascading failure process simulation is performed to explore the relationship between the degree of logistics nodes and the cascading failure of the complex logistics network under the linear/nonlinear load-capacity model.The simulation results show that with the increase of the number of iterations,the node survival rate curve of the nonlinear load-capacity model is better than the linear load-capacity model.The node survival rate curve of a logistics node with smaller degree under deliberate attack is more effective than a logistics node with larger degree.The research results show that the proposed nonlinear load-capacity model is more advantageous than the commonly used linear load-capacity model,and the greater the degree of logistics nodes being deliberately attacked in a complex logistics network,the more serious the consequences of cascading failure,which provides a new direction for the deliberate attack methods of the cascading failure phenomenon of complex logistics networks in this paper.(4)Analysis of cascading failure model of complex logistics network based on nonlinear load-capacity modelDeliberately attacking the logistics node with largest degree in the complex logistics network model,based on four complex logistics network cascading failure models under the linear/nonlinear load-capacity model(Initial load redistribution strategy cascading failure model under linear load-capacity model,differential load redistribution strategy cascading failure model under linear load-capacity model,initial load redistribution strategy under nonlinear load-capacity model and nonlinear load-capacity model under differential load redistribution strategy model),simulation analysis of cascading failure model of complex logistics network are performed,the node survival rate is set as evaluation index for network robustness.The simulation results show that based on the four models,when the value?is a fixed value,the node survival rate increases as the value?increases,when the value?increases,the critical threshold?increases and the network robustness is becoming worse,and based on the linear load-capacity model,the cascading failure simulation results are the same under the initial load redistribution strategy and the differential load redistribution strategy,which is consistent with the results derived from the previous theory,when the values?are the same,compared with the linear load-capacity model,the critical threshold?of the nonlinear load-capacity model is smaller,and the critical threshold of the differential load redistribution strategy model is the smallest under the nonlinear load-capacity model.That is,among the four cascading failure models of complex logistics networks,the complex logistics network cascading failure model based on the nonlinear load-capacity model with differencial load redistribution strategy is a better model,which is more robust.Based on the above optimal complex logistics network cascading failure model,controllability robustness_iP and network cost effectiveness e are used as performance evaluation indicators for complex logistics network for cascading failure,simulation of controllability robustness and economy of complex logistics networks is conducted.The simulation results show that controllability robustness_iP increases with increasing cost effectiveness e which means that increasing the construction cost of logistics network can improve the controllability robustness of logistics networks,controlling the capacity parameter?according to the specific requirements of the logistics cost-utility interval,we can optimize the control robustness of the logistics network.Under the premise of certain cost-effectiveness e,when 0.8???1.4,the larger capacity paramete?is,the stronger the controllbility robustness is.Especially,when0.8???1.4,the cost efficiency e of logistics can be selected in the section0.15?e?1.25according to the cost requirements of the logistics infrastructure.The research results show that the complex logistics network cascading failure model using the nonlinear load-capacity model and differential load redistribution strategy can more accurately and flexibly control the robustness and cost of the logistics network utility by adjusting the cost utility and capacity parameters of the complex logistics network,while reducing network costs and enhancing the ability of the network to resist attacks,this is of great significance to the robustness and economics of logistics networks.(5)Construct and analyze the dynamic cascading failure model of complex logistics networksConsidering the dynamic characteristics of real logistics node loads,we add dynamic load factors,a dynamic cascading failure model for complex logistics networks based on load capacity nonlinear relationships and differential load redistribution strategies is established for the first time.Based on the complex logistics network model,dynamic cascading failure model is compared with the non-dynamic cascading failure model.The simulation results show that the larger the capacity parameter?,the smaller the critical threshold?after cascading failure,the more robust the network is,and when the capacity parameter?is fixed,the five simulation results show that the critical threshold?of dynamic cascading failure models is less than the critical threshold?of the non-dynamic cascading failure model.Conclusively,the research method proposed in this paper is scientific and reasonable,and the dynamic cascading failure model of complex logistics networks is the optimal model.The complex logistics network after cascading failure is the most robust and can be used to provide a better solution.