本文探讨了用于产业结构和产业关联分析的图论模型的研制问题，并在投入产出模型的基础上建立了一类图论模型．应用实例表明，利用这类图论模型，我们可以确定产业的不同意义下的产业集，从新的角度分析主导产业的选择等，从而能在深层次上认识产业结构及产业关联状况． This article explores the development of graph theory models for industrial structure and industry correlation analysis, and establishes a class of graph theory models based on input-output models. The application example shows that by using this kind of graph theory model, we can determine the industrial clusters in different meanings of industries and analyze the choice of leading industries from a new perspective, so that we can understand the industrial structure and industrial relations at a deeper level.