基于多目标遗传算法的潜器外形优化设计

Optimization of Submersible Shape based on Multi-Objective Genetic Algorithm

计算流体力学软件应用于潜水器外形优化设计时,精度虽高,但效率低下。因此研究了试验设计和响应面模型技术,提出了一个潜水器外形优化设计的策略。其基本思想是:首先应用Gambit软件的日志文件建立起水动力分析模型,然后根据设计要求选取设计变量,用试验设计的方法在设计变量空间里选取样本点并进行阻力性能计算,得到各样本点的响应,并建立阻力、包络体积的二阶多项式模型响应面模型。潜水器设计时需要考虑能源与设备布置要求,因此将阻力与体积作为艇型优化的两个目标,研究了多目标遗传算法,并给出了Pareto最优解集。结果表明,文中采用近似模型的艇型优化过程,不但效率得到提高,精度也能得到保证,而且由CFD结果建立的阻力估算公式可以为后续设计带来很高的参考价值。

It is accurate to use CFD software in the optimization of submersible hydrodynamic shape,but its efficiency is low.Based on the statistical techniques,a strategy for optimization design of submersible shape is proposed.Its central idea is that：firstly the hydrodynamic model is set up by using the log file of Gambit,and design variables are chosen by referring design requirements;then the samples for analysis are created in the design space and finite element models corresponding to the samples are built and analyzed;thirdly the response surface models of drag and surrounded volume are constructed using these samples and responses obtained by hydrodynamic calculation.During the design of submersible shape,both drag performance and arrangement condition are factors to be considered.So multi-objective genetic algorithm is studied in this paper to solve the shape optimization with two objectives which are minimum drag and maximal surrounded volume,and the Pareto optimal solutions are given in the end.The result shows that this method can improve the efficiency and achieve optimal intention and the approximation of drag has valuable reference information for later research.