在群体遗传结构及规律的基础上，提出了一种改进的遗传算法——基于域进化模型的遗传算法。该算法模拟自然界中生物多子群的进化结构，定义三种新的算子——迁徒、正聚类和反聚类来描述子群间个体的随机或非随机移动，以此来改善遗传算法的性能；同时，通过引入模糊环境因子来设置各子群环境状况，并确定遗传算法的控制参数，以区分不同子群的遗传规律。针对多峰函数优化问题的仿真实验表明该算法具有较强的寻求全局最优点及避免过早收敛的能力，并将该算法应用于求解原油蒸馏过程优化问题 Based on the genetic structure and regularity of population, this paper proposes an improved genetic algorithm - genetic algorithm based on domain evolutionary model. The algorithm simulates the evolutionary structure of biological multiple subgroups in nature and defines three new operators - migration, normal clustering and anti-clustering to describe the random or non-random movement of individuals between subgroups in order to improve the genetic At the same time, the environmental conditions of each subgroup are set by introducing fuzzy environmental factors, and the control parameters of the genetic algorithm are determined to distinguish the genetic laws of different subgroups. Simulation experiments on the multi-peak function optimization problem show that the algorithm has a strong ability to seek the global optimum and avoid premature convergence. The algorithm is applied to the optimization of the crude oil distillation process