基于组合优化算法的结构稳健优化设计方法研究与应用

发布时间：2018-10-25 11:14

【摘要】：在传统的结构优化设计研究中,设计变量通常被视为确定的量,并没有考虑它们可能产生的的不确定性,这也最终会导致结构性能的不稳定。因此有必要考虑设计变量实际中存在的变差,本文即以火炮中的典型部件为研究对象,研究和探索了结构稳健优化设计方法在典型机械结构中的应用。首先对结构稳健优化设计方法进行了深入的研究,在iSIGHT中进行了优化算法的二次开发,将多岛遗传算法与改进Hooke-Jeeves算法相结合,提出一种基于组合优化算法的6σ结构稳健优化设计方法。其次建立了闭气结构和后大架的三维实体参数化模型,并在ABAQUS里进行静力学刚强度分析,评估其结构性能。在iSIGHT中,搭建结构稳健优化设计平台,集成PRO/E和ABAQUS,实现机械结构的“设计—分析—稳健优化—再设计”。最后提出一种中心加密试验设计方法,对初始设计空间进行更加有效的探索；为了得到设计响应和设计变量之间的关系,需要进行大量的实验,这必将消耗大量的资源和时间,因而本文分析研究了响应面与Kriging两种近似模型,将Kriging近似模型应用到实例中,极大地降低了计算时间；多岛遗传算法中的迁移操作保持了优化解的多样性,提高了包含全局最优解的机会,但收敛速度较慢,本文开发的改进Hooke-Jeeves算法,可以进行有效的局部加速搜索,提高了得到全局最优解的机会,应用本文开发的这种组合优化算法,可以很好的提高结构稳健优化设计的效率。闭气结构与后大架稳健优化设计的成功,说明了本文研究与探索的结构稳健优化设计方法是具有一定的实用价值,可以解决机械结构设计中的一些实际问题。
[Abstract]:In the traditional research of structural optimization design, design variables are usually regarded as deterministic quantities, and the uncertainties that they may produce are not considered, which will eventually lead to instability of structural performance. Therefore, it is necessary to consider the variation of design variables in practice. In this paper, the application of structural robust optimization design method in typical mechanical structures is studied and explored, taking the typical parts of artillery as the research object. Firstly, the structure robust optimization method is deeply studied, and the optimization algorithm is redeveloped in iSIGHT. The multi-island genetic algorithm is combined with the improved Hooke-Jeeves algorithm. A 6 蟽 structure robust optimal design method based on combinatorial optimization algorithm is proposed. Secondly, the three-dimensional solid parameterized model of the air-closed structure and the rear frame is established, and the statically rigid strength analysis is carried out in ABAQUS to evaluate its structural performance. In iSIGHT, the structural robust optimization design platform is built, and PRO/E and ABAQUS, are integrated to realize the "design-analysis-robust optimization-re-design" of mechanical structures. Finally, a central encryption experiment design method is proposed to explore the initial design space more effectively. In order to obtain the relationship between design response and design variables, a large number of experiments are needed, which will consume a lot of resources and time. Therefore, the response surface and Kriging approximation models are analyzed and studied in this paper, and the Kriging approximation model is applied to an example, which greatly reduces the computing time, and the migration operations in the multi-island genetic algorithm maintain the diversity of the optimal solutions. The opportunity to include the global optimal solution is improved, but the convergence rate is slow. The improved Hooke-Jeeves algorithm developed in this paper can carry out effective local acceleration search and improve the chance of obtaining the global optimal solution. It can improve the efficiency of structural robust optimization design. The success of the structural robust optimization design of the closed air structure and the rear frame shows that the structural robust optimization design method studied and explored in this paper is of certain practical value and can solve some practical problems in the mechanical structure design.
【学位授予单位】：南京理工大学
【学位级别】：硕士
【学位授予年份】：2011
【分类号】：TH122

【参考文献】