基于复合形与神经网络的机翼结构优化设计
详细信息 本馆镜像全文    |  推荐本文 | | 获取馆网全文
摘要
从元件位置和元件尺寸两个方面考虑,提出了一种全局结构优化设计方法。利用NASTRAN完成尺寸优化,利用神经网络对NASTRAN优化结果进行映射,用复合形法进行位置寻优,最终完成了机翼结构的优化。算例表明利用神经网络来预测NASTRAN优化结果有较高的精度与适用性,本文所提方法可行、正确,具有很高的效率。
From the thinking of both position variables and size variables,a global method that can solve the optimization problems was present in this paper.Letting the NASTRAN optimize the size variables,using the neural network to predict the calculated result of the NASTRAN,and this paper went on to optimize the position variables with the complex method.The example shows the structural optimization method that use Neural Networks to predict the result of the NASTRAN has high precision and adaptability.this method is not only correct and feasible,but highly effective and practically convenient.
引文
[1]Ji Yao Shen.Frederick Ferguson.Weight Reduction Through Optimal Arrangement of Force-Carrying Components Within Wing Box Structures[J].AIAA2004-4495.
    [2]Keunedy M P,Chua L O.Neural networks for nonlinear programming[J].IEEE Trans.Circuits Systems,1998.
    [3]邓扬晨等.基于分级优化的飞机翼面结构布局综合技术研究[J].强度与环境,2005,32(1):28~37.
    [4]王小妮.基于PATRAN环境下的参数化建模技术[J].飞机工程,2004,(4):26~29.
    [5]李文勇,李泉永.利用神经网络进行近似分析的结构优化设计[J].广西科学,2001,8(2):86~89.
    [6]欧海龙,李向真.多层复合形法在结构优化中的程序设计[J].世界地震工程,2000,16(14):113~117.
    [7]胡光良,李思明.神经网络在结构优化设计中的应用[J].工程建设与设计,2003,(6):16~17.
    [8]闻新等.MATLAB神经网络仿真与应用[M].北京:科学出版社,2003.

版权所有:© 2023 中国地质图书馆 中国地质调查局地学文献中心