Multi-objective optimal approach for injection molding based on surrogate model and particle swarm optimization algorithm

An integrated optimization strategy based on Kriging model and multi-objective particle swarm optimization(PSO) algorithm was constructed.As a new surrogate model technology,Kriging model has better fitting precision for nonlinear problem.The Kriging model was adopted to replace computer aided engineering(CAE) simulation as fitness function of multi-objective PSO algorithm,and the computation cost can be reduced greatly.By introducing multi-objective handling mechanism of crowding distance and mutation oper...     

Structural optimization of a circumferential friction disk brake with consideration of thermoelastic instability

This research suggests a new disk brake design using circumferential friction on the disk of a front-wheel-drive passenger car. The paper compares mechanical performance between the conventional and suggested disk brakes under dynamic braking conditions. Thermoelastic instability is considered in simulation of the test condition. An optimization technique using a metamodel is introduced to minimize the weight of the suggested disk brake. To achieve this goal, the response defined in the optimization formulation is expressed in a mathematically explicit form with respect to the design variables by using a kriging surrogate model, resulting in a simple optimization problem. Then, the simulated annealing algorithm is utilized to find the global optimum. The design results obtained by the kriging method are compared with those obtained from ANSYS analysis.     

基于Kriging模型的武器装备体系效能分析

为了分析影响装备体系作战效能的各个输入因素的变化对作战效能的影响,采用一种基于方差的敏感性分析方法——Sobol指数法,对体系模型的数据进行分析;采用基于Kriging模型的方法对Sobol指数进行计算。最后运用该方法对一防空装备体系数据进行分析,确定了影响作战效能的关键因素,为进一步的体系效能优化提供了依据。     

高效优化算法在船舶力学中的应用研究

作为一种贝叶斯优化算法,高效全局优化算法(EGO)利用克里格模型来构造近似模型,并采用样本填充准则以寻找下一个样本点来更新近似模型.文中详细介绍了该优化算法,并将其应用于船舶力学的两个典型优化例子.其中一个是潜艇的多学科概念设计,考虑了水动力、推进、重量、性能和成本5个学科;另外一个是屈曲状态下加筋板的优化问题.与传统优化相比,高效全局优化算法不仅收敛到全局最优解,而且更加有效.结果表明高效优化算法非常适用于船舶力学中的优化问题.     

基于试验数据的大跨度拱桥有限元模型修正

为了获取菜园坝长江大桥的基准有限元模型,结合Kriging代理模型和一种改进的粒子群优化算法,利用荷载试验数据对其初始有限元模型进行修正。首先,叙述模型修正和Kriging模型基本理论,在基本粒子群算法中引入交叉变异计算,提出一种改进的粒子群算法,并通过测试函数对改进的粒子群算法进行验证;其次,简要介绍菜园坝长江大桥荷载试验、荷载试验结果及初始有限元模型;最后,根据敏感性分析选定6个待修正参数,通过试验设计得到频率和位移关于修正的参数的样本,并建立有限元模型的Kriging代理模型以预测结构响应;以频率和位移的试验值和计算值残差为目标函数,分别利用基本粒子群算法和改进的粒子群算法在修正参数的设计空间内寻找目标函数的最小值,并对比分析模型修正的结果。结果表明：测试函数表明改进的粒子群算法具有较好的稳定性和成功率,并能获得更为精确的优化结果;建立的Kriging代理模型均方根误差较小,可以替代有限元模型预测结构频率和位移;经过模型修正,菜园坝长江大桥前5阶频率计算值与试验值相对误差均控制在5%之内;除个别测点外,位移相对误差均控制在10%以内;相比基本粒子群算法,改进的粒子群算法获得了更小的目标函数值,修正后的频率和位移的相对误差更小。     

汽车声学包轻量化设计

本文旨在研究汽车声学包的轻量化设计,要求在保证声学性能的前提下,尽可能实现轻量化的目标.首先运用统计能量法建立了包含车身结构和各声腔的整车模型,提取80 km/h匀速工况下发动机舱的声辐射激励、动力总成激励和车身表面脉动压力激励,并将其施加在该模型上,获得驾驶员头部声腔声压,结果与试验数据吻合良好,验证了模型的有效性....     

基于碰撞安全性的铝合金吸能盒轻量优化

本文中以一种匹配碳纤维复合材料保险杠防撞梁的吸能盒为研究对象,通过试验设计优化,获得吸能盒最佳锥角和溃缩孔的直径及其排列;然后采用拉丁超立方抽样和基于Kriging代理模型加点策略的多目标粒子群优化算法对吸能盒厚度进行多目标优化.优化后不仅吸能盒吸能量大幅提升,且吸能盒质量明显下降.刚性壁100％重叠率正面碰撞和25％...     

基于Kriging模型的车身耐撞性优化设计

采用最优拉丁方法进行试验设计,基于Kriging模型建立汽车耐撞性评价指标的替代模型,然后利用替代模型建立耐撞性优化的数学模型并进行耐撞性优化。结果表明,基于Kriging的替代模型的拟合精度较高,采用该模型进行的优化可提高汽车的耐撞性。     

基于代理模型的水下航行体头型优化设计方法研究

文章采用Kriging代理模型技术针对水下航行体头型优化设计开展研究。采用CFD方法分析了不同头型水下航行体的流体动力性能,将性能参数作为初始样本建立代理模型;根据测试样本点的预测均方误差选择加点策略更新代理模型,提高代理模型预测精度;采用非劣分类遗传算法对阻力和表面压力开展多目标优化,寻找近似最优解。结果表明:基于代理模型技术的水下航行体头型多目标优化设计方法可以有效提高设计效率,获得具有良好水动特性的航行体头型。     

Bi-fidelity Kriging model for reliability analysis of the ultimate strength of stiffened panels

A method based on a Bi-fidelity Kriging model is proposed for structural reliability analysis. It is based on adding low-fidelity data samples to the model to predict high-fidelity values, thus saving computational effort. Distance Correlation develops the correlation between the low and high-fidelity functions, initially proposed to assess the correlation between two variables. The bi-fidelity Kriging response surface model's efficiency as a surrogate model will be assessed for structural reliability problems that demand high computational costs, such as nonlinear finite element analysis structural models. The efficiency assessment is performed by comparing the accuracy of the failure probability predictions based on the Subset Simulation and First-order reliability method using the Bi-fidelity Kriging model as a surrogate for the performance function. The idea is illustrated by considering a representative component of marine structures analyzed by finite element analysis to create bi-fidelity scenarios to assess structural reliability with many variables. The results show that the proposed multi-fidelity method can provide an accurate failure probability estimation with less computational cost.