基于遗传算法的多级固体火箭总体/发动机一体化设计优化研究  

Integral Design Optimization for Multistage Solid Rocket and It's Motor Based on Genetic Algorithm

作　　者：孙丕忠[1] 夏智勋[1] 黄琳[1] 

机构地区：[1]国防科技大学航天与材料工程学院,长沙410073

出　　处：《宇航学报》2005年第B10期1-4,15,共5页Journal of Astronautics

摘　　要：在综合考虑发动机设计、火箭气动特性和外弹道关系的情况下,建立了多级固体火箭总体发动机一体化设计优化模型。该模型具有如下特点:一是不能保证目标函数和约束是凸集上的凸函数,存在“多峰”现象,二是设计变量含连续/离散混合设计变量。根据上述特点,优化方法选用遗传算法,并对遗传算法进行了改进研究,给出了一种改进的基于方向的变异算子。应用该算法完成了某三级固体火箭总体/发动机十五个设计参数的优选,取得了满意结果。Considering of SRM performance,rocket aerodynamics and trajectory, A set of mathematical models for multistage solid rocket design optimization were developed. The models have two characters：one with nonconvex design space, the other with continuous/discrete design variables. To find the global optimum solution of the problems, genetic algorithm was adopted. An improved mutation operator based on gradient direction was presented. The optimizing calculations with 15 variables for a third stage solid rocket were finished and the results are satisfactory.

关 键 词：遗传算法 固体火箭 一体化设计 优化 

分 类 号：V430[航空宇航科学与技术—航空宇航推进理论与工程]