A feasible approach to handle many constraints optimization problems based on Gaussian process frame

来源 :Asian Congress of Structural and Multidisciplinary Optimizat | 被引量 : 0次 | 上传用户：gdat86

【摘要】

：

【作者】

：

Wei Hu Hu Wang

【机构】

：

State Key Laboratory of Advanced Design and Manufacturing for Vehicle Body, Hunan University, Changs

【出处】

：

Asian Congress of Structural and Multidisciplinary Optimizat

【发表日期】

：

2018年11期

【关键词】

：

BPNN efficient global optimization (EGO) constrained optimization infeasible ini

下载到本地 , 更方便阅读

下载此文赞助VIP

声明 : 本文档内容版权归属内容提供方 , 如果您对本文有版权争议 , 可与客服联系进行内容授权或下架

论文部分内容阅读

　　Surrogate-assisted optimization algorithm (SAO) is an efficient way to save the time and cost of the evaluation for engineering optimization problems.Efficient Global Optimization (EGO)is an important SAO algorithm which is based on GP process and Expected Improvement (EI)criterion (Jones and Schonlau 1998).This algorithm was originally proposed to solve unconstrained problems and then was modified to solve the constrained problems (Bagheri and Konen 2017).

其他文献

AA1060铝箔/Q235钢的高能量利用率爆炸焊接

针对箔材的爆炸焊接,提出了一种新的装药结构,即在蜂窝结构炸药的基础上,用胶体水对炸药上端自由面进行覆盖.该装药结构减小了空中稀疏波对爆轰的弱化,从而可以显著地降低炸药的临界厚度.本文使用5 mm厚的低能乳化炸药作为爆炸材料,0.2 mm厚的铝箔和20 mm厚的钢板作为复板和基板,进行了爆炸焊接实验.理论计算了新装药结构下的爆炸焊接动态参数,并结合爆炸窗口对焊接质量进行了预测.

会议

爆炸焊接箔材能量利用率焊接窗口结合质量

聚脲弹性体涂层抗高速破片侵彻性能试验研究

研究了高速破片侵彻下，聚脲弹性体涂层对涂覆钢板结构抗侵彻性能的影响作用.通过弹道试验加载3.3 g立方体破片以1300 m/s以上速度撞击相同面密度的3.5mm厚无涂层钢板与含6mm涂层的2.7 mm厚钢板，含涂层钢板包括迎弹面涂层与背弹面涂层两种涂覆方式，获得破片贯穿不同类型靶板后的剩余速度，分析了靶板的能量吸收与抗侵彻性能.结果表明：相同面密度下，聚脲弹性体涂层能够提高涂覆钢板结构的抗侵彻性

会议

聚脲弹性体立方体破片高速侵彻涂覆钢板

Optimal design of product aesthetics considering variation of customers' kanesi

会议

Conservative Reliability Index for Non-Gaussian Distributions in Reliability-Based Design Optimizati

This study focuses on developing an uncertainty quantification method for reliability-based design optimization (RBDO) when both aleatory and epistemic uncertainty present.Recently,several efficient u

会议

Design of Auxetic Metamaterial-Inspired Piezoelectric Energy Harvesting Devices under Uncertainty

Energy harvesting that converts ambient,otherwise wasted,energy sources into usable electricity,has received considerable attention to power wireless sensors.Among various energy harvesting technologi

会议

Statistical model calibration using bias ratio to consider model form uncertainty:A case study of an

Due to favorable cost/benefit analysis,Computational Aided Engineering (CAE) techniques have become a nearly ubiquitous step in the engineered product development process.To improve the accuracy of CA

会议

CNN-Based RNN for Time Serial Modeling Problems

In this study,we hope to construct a time serial model for time-dependent problems such as high velocity impact or stretching by using recurrent neural network based on convolutional neural network (R

会议

RNNCNNtime serial modeltime-dependent problem

Design optimization of variable blank holder force and slide velocity for high productivity in sheet

会议

An improved LHS approach for constrained design space based on successive local enumeration algorith

Computational simulations for engineering design and optimization are time-consuming and difficult to obtain feasible solutions.Given that,metamodels have been proposed and widely used as substitutes

会议

design of experiments(DoE)LHSsuccessive local enumerationconstrained design s

A Strategy for Adaptive Sampling When Sampling Cost Is Variable in Design Space

会议

A feasible approach to handle many constraints optimization problems based on Gaussian process frame

与本文相关的学术论文