Study on optimization of collaborative innovation network structures based on grey property of knowledge flows

This paper establishes a model of collaborative innovation based on grey system theory and explores how the efficiency of innovation is influenced by the structures of collaborative innovation networks. The result shows that when the degree of grayscale of knowledge flow is low, the optimistic collaborative innovation network structure is random network which is characterized by low average path length; when the degree of grayscale of knowledge flow is medium, the optimistic collaborative innovation network structure is small world network which is characterized by high small-world coefficient; when the degree of grayscale of knowledge flow is high, the optimistic collaborative innovation network structure is regular network which is characterized by high average clustering coefficient. Two reasons of the results are derived. First, when the degree of grayscale of knowledge is low, it is easy for innovators to obtain certain kinds of knowledge, so expanding knowledge variety becomes more important for innovation; on the contrary, obtaining certain kinds of knowledge is more important for innovation. Second, high average clustering coefficient of the network is good for obtaining certain kinds of knowledge; low average path length is good for the diversification of knowledge. At last, some advices are proposed.