Reverse Logistics Network Modeling And Robust Study Based On Complex Network Theory

With the rapid advance of social reforms, citizens’ living standard has improved steadily. In the meanwhile, the development of economy also causes serious problems like resource depletion and environmental pollution. Reverse logistics not only can avoid waste of resources, reduce environmental pollution, but also cut down enterprise cost, improve citizens’ satisfaction of the enterprise and government. So reverse logistics has raised highly concern from the government, enterprises and citizens. Reverse logistics network is the foundation of reverse logistics operation. Nowadays, quite a lot scholars have studied reverse logistics. And most of them consider local network as the object of their studies. For example, the optimization of path location in the reverse logistics network, network node planning and so on. Few studies is carried out from the point of dynamic evolution of the entire system. This thesis establishes reverse logistics network evolutionary model based on complex network theory. The simulation of the model is done to study the network characteristic parameter and robustness and analysis the operation mechanism of reverse logistics network.First of all, this thesis summarizes the type, structure characteristics, complex network evolutionary model and statistic characteristics of reverse logistics. Analysis on the factors that can influence the reverse logistics in the process of network evolution is also carried out. And the network evolution is described by three stages separately. The generation of the complexity in the reverse logistics system evolution is put forward and the characteristics of the complexity is also pointed out.Secondly, reverse logistics network evolutionary model of hierarchical node local selection is built. Five kinds of nodes is set up in the reverse logistics:recycle bin, recycle center, disposal center, re-manufacturing center and reuse market. To reflect the evolution process of the network, a variety of changes in the network structure is set up. The connection range of the selected nodes is restricted to the local world of the hierarchical network, and it is picked up through the selection of merit. Through model simulation, network topology can be got to calculate node degree distribution, average path length, and node clustering coefficient distribution. The result shows that reverse logistics network follows the power law distribution.Last, random attack and target attack is designed to study the robustness of the reverse logistics when under disturbance of external factors. And the node degree distribution, network efficiency and maximum network cluster coefficient is also analyzed under disturbance. To ensuring the normal operation of the network, suggestions of risk aversion and risk recovery are proposed to increase the stability of the network.