为了解决复杂工程优化问题中计算效率与代理模型精度之间的矛盾，本文基于最大最小距离准则和最小化代理模型预测准则提出一种多点加点准则，构建改进序列近似优化方法的算法流程。以二维Rosenbrock函数和Golinski减速器优化问题为例，与传统方法相比，改进方法得到的全局最优解精度更高，所需构造样本更小。将该方法应用到某型水下航行器水动力外形优化问题中，以巡航阻力系数和空间损失率最小为目标，建立优化问题的数学模型，原始计算模型仅调用80次之后优化收敛，大大提高了设计效率，优化方案巡航阻力系数和空间损失率比初始方案分别降低了9.81%和0.28%。研究表明，改进序列近似优化方法可显著提高代理模型精度，相同计算条件和时间下，具有更高的优化效率和优化精度。
In order to balance the contradiction between computational efficiency and the accuracy of surrogate models in complicated engineering optimization, a multi-point infilling criterion was proposed based on maximin distance criterion and minimizing the predictor criterion. And the algorithm flow process of enhanced sequential approximate optimization was constructed. The numerical examples of two-dimensional Rosenbrock function and Golinski reducer optimization were carried out, compared with the traditional method, the global optimal solution obtained by the enhanced method has higher accuracy and requires smaller samples. Then the method is applied to a underwater vehicle hydrodynamic shape optimization. The mathematical model of the optimization is established with the goal of minimizing the drag coefficient and the space loss rate.The optimization simflow converged after 80 times of original model calling. The design efficiency was increased greatly and the resistance coefficient and space loss rate of the optimized scheme are respectively 9.81% and 0.28% less than primary scheme. The research shows that the proposed method can significantly improve the accuracy of the surrogate model. Under the same computing condition and time, it has higher optimization efficiency and accuracy.
2022,44(2): 121-128 收稿日期：2021-01-18
DOI：10.3404/j.issn.1672-7649.2022.02.022
分类号：TP242.3
基金项目：中国科学院前沿基础研究项目(QYJC201913)；十三五预研项目(2020107/2002)
作者简介：高伟(1996-)，男，硕士研究生，从事水下机器人多学科优化方法研究
参考文献：
[1] 宋保维, 王新晶, 王鹏. 基于变保真度模型的AUV流体动力参数预测[J]. 机械工程学报, 2017, 53(18): 176–182
[2] HAN Z H, ZHANG K S. Surrogate-based optimization[M]// Real-World Applications of Genetic Algorithms. InTech, 2012.
[3] HAN Z H, GOERTZ S. Hierarchical kriging model for variable-fidelity surrogate modeling[J]. Aiaa Journal, 2012, 50(9): 1885–1896
[4] 黄礼铿, 高正红, 张德虎. 基于变可信度代理模型的气动优化[J]. 空气动力学学报, 2013, 31(6): 783–788
[5] 刘蔚. 多学科设计优化方法在7 000米载人潜水器总体设计中的应用[D]. 上海: 上海交通大学, 2007.
[6] NAKAYAMA, HIROTAKA, YUN, et al. Sequential approximate multiobjective optimization using computational Intelligence[M]. Springer Berlin Heidelberg, 2009.
[7] KITAYAMA S, ARAKAWA M, YAMAZAKI K. Sequential approximate optimization using radial basis function network for engineering optimization[J]. Optimization and Engineering, 2011, 12(4): 535–557
[8] WANG D, WU Z, FEI Y, et al. Structural design employing a sequential approximation optimization approach[J]. Computers & Structures, 2014, 134(apr.): 75–87
[9] JACOBS J H, ETMAN L F P, KEULEN F V, et al. Framework for sequential approximate optimization[J]. Structural and Multidisciplinary Optimization, 2004, 27(5): 384–400
[10] 谢延敏, 张飞, 潘贝贝, 等. 基于并行加点kriging模型的拉延筋优化[J]. 机械工程学报, 2019, 55(8): 73–79
[11] 柳斐, 韩忠华, 张瑜, 等. 一种基于多点EI的高效全局气动优化设计方法[C]// 2017年(第三届)中国航空科学技术大会论文集(上册). 2017.
[12] VIANA F A C, HAFTKA R T, WATSON L T. Efficient global optimization algorithm assisted by multiple surrogate techniques[J]. Journal of Global Optimization, 2013.
[13] 陈国栋, 卜继玲. 基于序列径向基函数的多目标优化方法及其应用[J]. 汽车工程, 2015, 37(9): 1077–1083
[14] WANG D, XIE C. An efficient hybrid sequential approximate optimization method for problems with computationally expensive objective and constraints[J]. Engineering with Computers, 2020(4).
[15] 彭科, 胡凡, 张为华, 等. 序列近似优化方法及其在火箭外形快速设计中的应用[J]. 国防科技大学学报, 2016, 38(1): 129–136
[16] 傅珏生, 汪仁官. 最大最小距离的性质及计算[J]. 应用概率统计, 1999, 15(2): 193–198
FU Y S, WANG R G. The prorcrtics and computation of maxinmum-minimum distance[J]. Chinese Journal of applied probability and statistics, 1999, 15(2): 193–198
[17] LIU J, HAN Z H, SONG N P. Comparison of infill sampling criteria in kriging-based aerodynamic optimization[C]// 28th Congress of the International Council of the Aeronautical Sciences, 2012.
[18] ROSENBROCK H H. An automatic method for finding the greatest or least value of a function[J]. The Computer Journal, 1960, 3(3): 175–184
[19] PADULA S L, ALEXANDROV N, GREEN L L. MDO test suite at NASA langley research center[C]//Proceeding of 6th AIAA/NASA/ISSMO Symposium on Multidisciplinary Analysis and Optimization, Bellevue, 1996.
[20] LUO Y Z, TANG G J. Parallel simulated annealing using simplex method[J]. AIAA Journal, 2006, 44(12): 3143–3146
[21] 杨雪榕, 梁加红, 陈凌, 等. 多邻域改进粒子群算法[J]. 系统工程与电子技术, 2010, 32(11): 2453–2458
[22] 张铁栋, 姜大鹏, 盛明伟. 无人无缆潜水器技术[M]. 上海: 上海交通大学出版社, 2018.12.
[23] 王建. 多学科优化设计在水下无人航行器设计中的应用研究[D]. 哈尔滨: 哈尔滨工程大学, 2016.
[24] GAO W, GU H T, SUN Y, et al. Multi-objective design optimization of main body of an unpowered underwater vehicle based on surrogate models[C]//New York: IEEE, 2020: 791−796.
