Supply Chain Modeling And Performance Analysis Based On Complex Networks

Supply chains are complex socio-economic systems. Due to their inherent complexity, dynamics as well as uncertainties, it is hard to model supply chains and analyze their performances efficiently. The emergence of the complex network theory provides a novel approach to explore supply chains. Complex networks theory concentrates on investigating the mechanisms, which determine the topological properties observed in numerous complex systems, and exploring the relevance between these topological properties and the system behaviors. The main aim of this dissertation is to model supply chains and analyze their behaviors based on complex network theory. A structural evolving model for supply chain is proposed. Then two important behaviors, i.e. robustness to supply disruption and cooperation behaviors are investigated from the perspective of topological strucutures. The main results and contributions of this dissertation are as follows.1. Propose a structural evolving model, which focuses on the structural characteristics in supply chains. This model incorporates three mechanisms, i.e., growth, attachment preference and local selection. Then some important statistical properties of the model are explored, which include average path length, clustering coefficient, degree distribution, structural heterogeneity and degree correlation. The simulation results show that the model follows"small-world" effect and its clustering coefficient is extremely small; the model’s structure is heterogeneous and its degree distribution follow power-law distribution; the model is disassortative network. These model’s properties are consistent with previous works and empirical research, indicating that the model can well describe the structures in supply chain.2. Based on the structural evolving model, the robustness of supply chains is explored. The robustness in this dissertation is further divided into two categories: static and dynamic robustness. In the static analysis, four rules are introduced to simulate the different damages to supply chain, which include randomly deleting nodes, intentionally deleting nodes, randomly deleting links and intentionally deleting links. Considering nodes’capacity constraints, the cascading failures resulting from nodes’ overloading are investigated in the dynamic analysis. In order to imitate cascading failures, a failure propagation process is proposed. Results show that supply chains exhibit different robustness to different damages and the failures propagation can result in supply chains extremely fragile to supply disruption.3. Based on the structural evolving model, supply chains’cooperation behavior is studied. The strategy evolution of entities in supply chains is explored in the framework of evolutionary game and then two different games, prisoner’s dilemma game and snowdrift game, are respectively applied to the strategy evolution. The results indicate that structural heterogeneity, competition pressure and strategy payoff are all involved with cooperation levels in supply chain. High structural heterogeneity and strong competition pressure are helpful to promote cooperation frequency.