为了刻划进化过程中的种群结构,需要寻找匹配于优良解集的最佳贝叶斯网络图,种群中的染色体基因位置对应于贝叶斯网络图的顶点。本文研究了基于BD度量(Bayesian Dirichlet metric)的局部网络图结构,主要分析了存在一条从顶点a到c连接的局部结构B_(a→c)的度量,与从两个顶点a、b到c均有连接的局部结构B_(a,b→c)的度量的关系。 In order to characterize the population structure during the evolutionary process, we need to find the best Bayesian network map that matches the good solution set, and the chromosomal location in the population corresponds to the vertex of the Bayesian network graph. In this paper, we study the local network graph structure based on the Bayesian Dirichlet metric. We mainly analyze the existence of a measure of the local structure B_ (a → c) connected from vertices a to c, There is a relation between the measures of the local structure B_ (a, b → c) connected.