Complex networks provide a new tool for studying relationships among a large number of enterprises in industry. However, it is difficult for a single model to describe the competitive relationships between enterprises comprehensively and objectively. In order to solve this problem, we propose the model group theory in this paper. Taking the industry of Auto Parts in China as an example, the paper first models a bipartite network and a weighted bipartite network which sets the enterprises' sales volume of corresponding products as edge weight. Further, the paper constructs a Boolean competitive relationship network of auto Parts Manufacturers projected by the Boolean Bipartite Network and a directed competitive pressure network of Auto Parts Manufacturers projected by the Weighted Bipartite Network. Then the paper analyzes the relations and differences of the four models comprehensively and explains the economical backgrounds of the network models topological properties.