研究适于离散变量结构优化设计的遗传算法．探讨了离散变量结构优化问题的基因表达模式，提出了一种减小基因搜索范围的子空间构造方法．该算法可处理受应力、位移约束的结构优化问题．数值计算实例表明该算法效率高，具有很强的适应性 The genetic algorithm suitable for structural optimization of discrete variables is studied. The gene expression patterns of structural optimization problems with discrete variables were discussed, and a subspace construction method was proposed to reduce the scope of gene search. The algorithm can deal with structural optimization problems under stress and displacement constraints. Numerical examples show that the algorithm is highly efficient and adaptable