火箭炮底架多目标优化设计

Multi-objective Optimization Design of Under Frame of Rocket Launcher

针对某型火箭炮底架超重的问题,通过拓扑优化和多目标尺寸优化实现了底架的结构优化。基于多工况拓扑优化,寻找出在设计空间中最佳的材料布局。以关键零件厚度作为设计变量,采用哈默斯雷实验设计生成样本数据,使用径向基函数(RBF)神经网络模型建立底架质量、一阶模态频率、比刚度结构效能的代理模型。采用多目标遗传算法(MOGA)和序列二次规划法(SQP)的组合算法优化策略进行优化求解。结果表明,优化后底架质量减小15.2%,比刚度结构效能和一阶固有频率均得到改善,实现了底架轻量化目标。

Focusing on the overweight under frame of a rocket launcher,the structure optimization of the under frame was realized through topology optimization and multi-objective size optimization.The best distribution of under frame material was sought by muti-condition topology optimization.Taking the thickness of key parts as design variables,the sample data were generated by the design of a Hammersley experiment.A RBF neural network model was used to build the surrogate model of the under frame mass,first-order modal frequency and specific stiffness structure efficiency.The combinatorial algorithm optimization strategy of multi-objective genetic algorithm(MOGA)and sequential quadratic programming(SQP)was used to solve the optimization problem.The results showed that the mass of the optimized chassis was reduced by 15.2%;the structural efficiency of the specific stiffness and the first-order natural frequency were improved,the purpose of under frame weight reduction was achieved.