Study On Evolution Mechanism Of Supply Chain Network

With the unceasing progress of science and technology, the global economicintegration is speeding up and human economic activities are more and more complexand changeable. In such circumstances, enterprise’s management faces tremendouschallenges. In order to enhance enterprise’s ability to respond to the environmentalchanges, the supply chain management pattern has been burgeoning since1990s. Atpresent, the supply chain gradually constitutes a complex network, in which the nodesare linked dynamically via logistics, information flow as well as capital flow andinteract with each other.From the viewpoint of organization, supply chain network show significantecological attributes and constitutes an ecosystem formed by rational enterprises oflimited resources. In this system, enterprises with similar business can be regarded asindividuals of different populations. All the suppliers, manufacturers, distributors andretailers formed supplier population, manufacturer population, distributor populationand retailer population respectively, and these populations developed supply chaincommunities. This thesis conducts a qualitative analysis on evolution mechanism ofsupply chain network from the two levels of community and network system. The resultshows that the supply chain community displays characteristics of survival anddevelopment, with four dynamic evolution phases including birth, growth, maturation,self-renewal or recession and the network exhibits significant complex feature.From the perspective of complexity, the evolution mechanism of supply chainnetwork can be quantificationally analyzed based on the complex network theory. Thearticle analyzes the topology structure of the supply chain network and introduces apreferential attachment model with Poisson addition of new nodes and random deletionof existing nodes to describe its evolutionary process, considering the emergence,declining and withdrawal of the nodes in the real supply chain. After that, the degreedistribution of this model is discussed by continuum theory and simulation. The resultsshow that the model follows a power law degree distribution with degree exponent3, which is in line with the empirical data.As a part of the system, each node of the supply chain network has a special role inthe evolution process. Although an enterprise only directly affects finite nodes, thesphere of influence is far-reaching, especially the core enterprise. Therefore the theory related to dynamics is used to investigate the micro-mechanism of the supply chainnetwork via an empirical case. We take the whole data of purchase orders in a coreenterprise of a population as the empirical object. First, we analyze the purchase orderassignment and obtain a power-law distribution with exponent1, which implies therationality behind the preferential attachment mechanism of the above evolution model.Second, we capture the attributes of creation times of purchase orders to an individualvendor, as well as to all vendors, and further investigate whether they have some kind ofdynamics by applying logarithmic binning to the construction of distribution plot. It’sfound that the former displays a power-law distribution with approximate exponent2.0,while the latter is fitted by a mixture distribution with both power-law and exponentialcharacteristics. Obviously, the two distinctive distributions denote the bursty nature ofbusiness behavior in a supply chain network. To better describe the mechanismgenerating the heterogeneity of purchase order assignment process from the objectivecompany to all its vendors, a model driven by product life cycle is presented, and thenthe analytical distribution and the simulation result are obtained, which are in good linewith the empirical data.