Manufacturing enterprise collaboration network: An empirical research and an evolutionary model

With the increasingly fierce market competition, manufacturing enterprises have to continuously improve their competitiveness through their collaboration and labor division with each other, i.e., forming manufacturing enterprise collaborative network (MECN) through their collaboration and labor division is an effective guarantee for obtaining competitive advantages. To explore the topology and evolutionary process of MECN, this paper investigates an empirical MECN from the viewpoint of complex network theory, and constructs an evolutionary model reproduces the topological properties found in the empirical network. Firstly, a large-size of empirical data related to the automotive industry was collected to construct a MECN. Topological analysis indicates MECN is not a scale-free network, but is a small-world network with disassortativity. Small-world property indicates enterprises can respond quickly to the market, but disassortativity shows the risk spreading is fast and the coordinated operation is difficult. Then, an evolutionary model based on fitness preferential attachment and entropy-TOPSIS is proposed to capture the features of MECN. Besides, the evolutionary model is compared with a degree-based model which only considers node degree. The simulation results show the proposed evolutionary model can reproduce a number of critical topological properties of empirical MECN, while the degree-based model does not, which validates the effectiveness of the proposed evolutionary model.