基于组合代理模型的箱形梁轻量化设计  

Lightweight design of box girder based on combined agent model

作　　者：陈杨 秦义校[1] 杨俊乐 Chen Yang;Qin Yixiao;Yang Junle

机构地区：[1]太原科技大学机械工程学院,太原030024

出　　处：《起重运输机械》2023年第14期21-27,共7页Hoisting and Conveying Machinery

基　　金：山西省重点研发项目(201903D121067)。

摘　　要：文中针对常见的3种代理模型多项式响应面(RSM)、径向基(RBF)神经网络以及Kriging模型在结构优化中的代理效果进行对比研究。以桥式起重机箱形梁轻量化为例,分析了3种代理模型在确定系数下的代理效果,并以此为依据提出了1种适用于该优化案例的组合代理模型。最终,以智能算法与梯度算法的组合算法对得到的4种代理模型寻优,并将寻优结果通过有限元模型进行验证。实际工程案例表明,基于代理模型的结构优化算法,在保证优化效果的前提下,大幅度提高了优化效率。In this paper,the agent effects of three common proxy models,polynomial response surface(PRS),radial basis function(RBF)neural network and Kriging model,in structural optimization were compared.Taking the lightweight box beam of bridge crane as an example,the agent effects of three agent models under certain coefficient of determination were analyzed,and based on this,a combined agent model suitable for this optimization case was proposed.Finally,according to the combined algorithm of intelligent algorithm and gradient algorithm,the four kinds of agent models were optimized,and the optimization results were verified by finite element model.The actual engineering case shows that the structural optimization algorithm based on agent model can greatly improve the optimization efficiency on the premise of ensuring the optimization effect.

关 键 词：组合代理模型 确定系数 智能算法 箱形梁 布局优化 

分 类 号：TH215[机械工程—机械制造及自动化]