摘要
根据超空泡流动力学特性,将其流型计算处理为一有约束的优化问题,从而使空化器优化转化为两目标优化问题。结合基于精英保留策略与小生境技术,选择算子主体为并行选择法的改进混合遗传算法和绕空化器无粘流场的CFD分析进行优化求解,流场数值解采用非结构网格显式时间推进Jameson有限体积法解二维欧拉方程得到。基于该算法进行了以空化器阻力系数最小化为目标的优化设计,同文献结果的对比证明:该方法是有效的,并且极大地简化了计算。
Based on the special features of supercavitating flow,the computation of supercavitating regime was treated as a constrained optimization problem.Then the optimization of the cavitator could be translated into a two-object optimization.The improved hybrid genetic algorithm was proposed based on the elitism and niche technique,in which the main of selecting operators was the paralleling selection.The inviscosity flow field around the cavitator was solved by the numerical solution of two dimensional Euler equations using Jameson scheme on unstructured meshes with explicit time integration.Compared with the results in the references,the genetic algorithm was proved to be effective when used in the optimal design of the minimum pressure drag coefficient of the cavitator and meanwhile it could greatly reduce complexity in calculation.
出处
《海军工程大学学报》
CAS
北大核心
2010年第6期60-64,106,共6页
Journal of Naval University of Engineering
基金
国家部委基金资助项目(9140C300502070C30)
关键词
遗传算法
超空泡
多目标
空化器优化
genetic algorithm
supercavitating
multi-object
cavitator optimization