Inter-organization cooperation and innovation is increasingly common. However as the limitation of information, the members in the innovation network cannot evaluate each other when they need to choose cooperators. In this paper, a new type of inter-organization innovation network which comprises preferential attachment and preferential deletion based on grey relational analysis is studied. A series of simulations and analysis of average path length and clustering coefficient are conducted. To be specifically, grey relational analysis can help organizations in choosing cooperators. By simulation analysis, we compare the inter-organization innovation network in the steady state based on grey relational analysis with the random network and regular network, we find that such innovation network has the property of ‘small world’, which can demonstrate that evolution model of innovation network based on grey relational analysis is reasonable.