基于代理模型的高速列车齿轮箱裂纹识别

为建立裂纹结构动力响应与裂纹参数之间的解析关系从而对齿轮箱裂纹进行有效识别,提出一种可替代原有高精度分析模型的计算量小且计算精度较高的基于代理模型的裂纹识别方法。利用初始样本通过有限元与插值算法建立裂纹结构参数与动力响应之间的Kriging代理模型对应关系,从而代替原有的物理参数模型与结构响应关系,有效减少有限元计算次数,并通过随机粒子群优化方法对建立的代理模型进行全局裂纹参数寻优。通过一个悬臂梁结构的数值算例,对所提方法进行有效验证,并将该方法应用到某高速列车齿轮箱的裂纹识别中,结果表明该方法能够有效地对结构裂纹进行识别。     

基于RBFNN-ISSA的特大跨径悬索桥有限元模型修正

针对大跨径悬索桥一类复杂结构的有限元模型修正问题,提出了一种基于径向基神经网络(radial basis function neural network, RBFNN)子结构代理模型与改进麻雀搜索算法(improved sparrow search algorithm, ISSA)的有限元模型修正方法。首先,基于桥梁图纸数据采用通用有限元软件建立一座大跨悬索桥的初始有限元模型,并根据拉丁超立方抽样原则生成子结构材料参数-结构响应的训练样本,通过RBF神经网络和子结构模拟方法对初始有限元模型进行解构重组和样本学习,拟合关于材料参数-结构响应的代理模型。其次,建立考虑主梁挠度和模态频率误差最小的有限元模型参数修正数学优化模型,采用Tent混沌映射及黄金正弦策略改进标准麻雀搜索算法,引入柯西分布函数和贪心保留策略对每一代麻雀种群进行扰动,以用于求解联合静、动力特征的有限元模型修正数学优化问题。最后,以杭瑞高速洞庭湖大桥为工程背景,进行了悬索桥荷载试验,利用实测桥梁响应数据验证了该方法的可行性。研究结果表明：基于RBF神经网络与子结构法的模型修正方法,可以建立拟合精度较高的悬索桥结构代理模型;...     

基于响应面与灵敏度分析的区间不确定性参数识别方法

针对工程结构中普遍存在的不确定性,需开展考虑不确定性的有限元建模与模型参数识别。提出了具有区间不确定性参数识别的分步实施方法:①通过Box-Behnken矩阵设计方法进行响应面的样本点设计,并代入有限元模型计算获得结构的关心固有频率;基于二次多项式建立固有频率与待识别参数间的响应面模型;基于优化方法与自适应响应面思想对区间参数中值进行识别;②基于响应面模型并结合灵敏度分析,实现迭代区间参数半径的识别。通过某个质量—弹簧系统的数值算例和一组镜架系统工程实例的区间不确定性参数识别,验证了所提出方法的可行性和可靠性;参数识别结果表明所提出的方法具有较高的计算效率且可有效地避免区间优化导致的收敛问题。     

基于时域响应灵敏度分析的非线性系统参数识别

基于响应灵敏度分析的有限元模型修正法已广泛应用于线性结构系统的局部损伤和裂纹参数等的识别。尝试将该方法推广应用到非线性系统的参数识别中。从非线性系统运动方程出发,通过数值积分得到系统的强迫振动响应,对系统的物理参数求导得到时域响应对参数的灵敏度,从而构造相应的响应灵敏度矩阵用于参数识别反问题。以Holmes-Duffing非线性系统和物理工程中有着广泛应用的双重sine Gordon非线性系统的参数识别为例,说明方法的应用过程。算例研究了不同的初始参数和测量噪声对识别结果的影响,结果表明响应灵敏度法能准确快速识别非线性系统的参数,并且具有对测量噪声不敏感的优点。     

含裂纹轮对的振动特性分析及裂纹参数识别

为实现实际运行工况中的轮对裂纹特征提取及裂纹参数识别,利用Abaqus建立含裂纹轮对系统的有限元模型,仿真轨道随机不平顺激励下的轮对振动响应,并采用变分模态分解(VMD)和快速傅里叶变换(FFT)提取不同运行速度下的轮对裂纹特征参数.其次,利用Kriging代理模型构建裂纹参数和IMF分量中1X、2X谐波成分幅值的关系...     

基于贝叶斯优化BiLSTM模型的输电塔损伤识别

结构的加速度响应可以反映结构的状态信息，蕴含结构的损伤特征。针对目前输电塔健康监测系统产生大量数据而无法有效分析和诊断输电塔损伤的问题，利用结构输出加速度响应数据的时序关系，提出了基于双向长短时记忆网络(bi-directional long and short-term memory, BiLSTM)的损伤识别方法，并采用概率寻优方法贝叶斯优化(Bayesian optimization, BO)确定网络模型超参数。首先描述了BiLSTM的基本原理，给出基于贝叶斯优化的超参数选取策略，从而提出了基于BO-BiLSTM模型的损伤识别方法。然后使用该方法对输电塔有限元模型进行了损伤定位与模式识别，测试集的整体识别准确率达到94.2%。为了验证该方法对实际结构的损伤识别效果，提出基于异源数据的损伤识别方式：将输电塔有限元模型数据作为模型训练的样本训练BO-BiLSTM模型，使用试验数据用作验证集检验损伤识别效果。识别结果表明BO-BiLSTM可以较为准确的识别真实结构的损伤情况，识别效果较BiLSTM以及BO-LSTM更稳定。     

大型复杂钢管焊接结构损伤识别新方法

提出一种将Kriging代理模型和动力指纹方法相结合的大型复杂结构损伤识别两步法。首先建立结构响应与结构损伤参数之间的Kriging代理模型,代替结构原有有限元模型,对结构损伤进行全局识别,确定损伤位置;然后针对损伤结构,通过动力指纹方法,以结构单元模态应变能曲率差为指标,对结构损伤进行局部识别,确定损伤状况;最后,将该两步识别法应用于某导弹发射台模型的损伤识别中。实验结果表明,该方法可应用于工程实际,为大型复杂结构损伤识别提供了一种新思路。     

桥梁断面范德波尔振子涡激气动力模型参数非线性化的能量原理

采用范德波尔振子类涡激气动力模型，通过能量平衡原理，推导了涡激共振过程中结构的振幅增量、初始气动阻尼、非线性气动阻尼三者之间的基本关系，进而得出模型中气动力参数ε与振幅yT之间关系的识别原理。基于某扁平箱梁桥梁断面的节段模型涡振试验结果，对范德波尔类涡激气动力模型参数随振幅的演变关系进行了识别。结果表明，在涡激共振锁定区间内，随着振幅的增加，参数ε呈单调下降的趋势。与之相反的是，参数ε形成的非线性气动阻尼比却呈非线性增长的规律。当参数ε相关的非线性气动阻尼、初始气动阻尼、结构阻尼三者之和为零时，结构达到稳定的涡振极限环状态。研究表明：初始气动阻尼特性决定了结构能否起振而形成涡振锁定区间；识别出模型参数随振幅的变化关系后，高于试验阻尼的结构涡振响应具有可预测性。     

一种基于信息熵的分布参数结构传感器/激励器优化布置方法

针对用基于有限元方法的离散坐标体系研究传感器优化布置、所得仅为结构真实动力响应近似解直接影响传感器优化布置结果问题,提出基于贝叶斯统计系统识别方法与信息熵理论的分布参数结构传感器优化布置方法。以结构模型修正为目的,用贝叶斯统计系统识别方法识别结构模型参数最优值及不确定性程度,利用信息熵定量表征结构模型参数识别结果的不确定性程度;再将传感器优化布置问题转化为连续数值优化问题,以传感器位置为优化变量,通过遗传算法极小化模型参数识别结果的不确定性(即信息熵)识别传感器的最优布置位置;获得最大结构响应信息量,即识别结构模型参数的不确定性最小。通过双墩带弹性支撑的三跨连续梁桥数值仿真研究对该方法进行验证。     

基于VMD和DBN的非线性结构模型参数识别

为解决现有的非线性结构模型参数识别方法面临优化过程复杂的问题,提出一种基于变分模态分解(variational mode decomposition, VMD)和深度置信网络(deep belief network, DBN)的非线性结构模型参数识别方法。首先,利用VMD和希尔伯特变换(Hilbert transform, HT)识别振动响应的瞬时参数;将瞬时参数进行主成分分析后作为输入,非线性模型参数作为输出;然后,利用DBN拟合两者之间的非线性映射关系;最后,将实测振动响应的瞬时参数进行主成分分析,输入训练好的DBN可直接识别修正后的非线性模型参数。通过对两个不同非线性类型的双自由度模型和一个复杂框架模型在地震作用下的数值模拟,与高压输电结构的振动台试验,验证了该方法的有效性。数值与试验结果表明,所提方法具有较高的计算效率和良好的抗噪性。     

基于统计提升准则的注塑工艺参数优化

为了同时改善生产平板型注塑制品时的总体收缩度和收缩均匀度,提出基于统计提升准则的注塑成型工艺参数的多目标优化方法,寻找平衡两个质量指标的优化设计.首先利用小规模的实验设计方法获得建模数据集,针对应用中存在的建模数据奇异点问题提出一种数据预处理方法,并依此分别建立两个指标的初始替代模型,用于代替优化过程中代价高昂的计算分析;随后依据Pareto统计提升准则寻找新的采样点加入建模数据集来重新建模,使寻优结果不断趋近真实的Pareto前沿.仿真结果表明,较常规的建模优化方法,本文提出的方法能使用较少的采样数据,显著地改善平板制品的收缩质量.对于HDPE材质的矩形制品,保压曲线先恒定后线性递减可以获得好的收缩均匀度,使用压力上限值恒定保压可以获得好的平均收缩度.     

A framework for design optimization using surrogates

Design optimization is a computationally expensive process as it requires the assessment of numerous designs and each of such assessments may be based on expensive analyses (e.g. computational fluid dynamics method or finite element based method). One way to contain the computational time within affordable limits is to use computationally cheaper approximations (surrogates) in lieu of the actual analyses during the course of optimization. This article introduces a framework for design optimization using surrogates. The framework is built upon a stochastic, zero-order, population-based optimization algorithm, which is embedded with a modified elitism scheme, to ensure convergence in the actual function space. The accuracy of the surrogate model is maintained via periodic retraining and the number of data points required to create the surrogate model is identified by a k-means clustering algorithm. A comparison is provided between different surrogate models (Kriging, radial basis functions (Exact and Fixed) and Cokriging) using a number of mathematical test functions and engineering design optimization problems. The results clearly indicate that for a given fixed number of actual function evaluations, the surrogate assisted optimization model consistently performs better than a pure optimization model using actual function evaluations.     

Identification of multiple cracks based on optimization methods using harmonic elastic waves

A theoretical study of the identification of multiple cracks in an elastic medium based on optimization methods using harmonic elastic waves is presented in the paper. A general interacting crack model is first used to determine the dynamic interaction between arbitrarily located and oriented cracks subjected to plane harmonic waves. The solution of this problem is then implemented into an optimization process for the identification of unknown cracks from known strain components at discrete locations. An optimization scheme based on sensitivity analysis is used to determine the length, the orientation and the position of cracks. Numerical simulation indicates that the sensitivity analysis and the optimization method used are effective in identifying multiple cracks. It is observed that convergent results can be achieved from a set of arbitrarily determined initial values of the crack parameters. Numerical examples are presented to illustrate the determination of the length, orientation, and position of different interacting cracks.     

Operation parameters multi-objective optimization method of large vertical mill based on CFD-DPM

The association mechanism between the main operation parameters and multi-physical fields of the large-scale vertical mill system is unclear, which leads to the difficulty in optimizing operation parameters to improve the performance of large vertical mill systems. To investigate the mechanism of multi-physical field coupling in the operation of the large vertical mill, the numerical simulation method is constructed by coupled CFD-DPM model to calculate the finished product quality, the simulation results were in good agreement with the actual operation results. Based on the Kriging surrogate model, a multi-objective optimization framework for large vertical mills is proposed. Finally, the multi-objective optimization design of LGM large vertical mills is carried out. Combined with CFD-DPM coupling method is developed, design variables and output responses are determined. The Kriging method is used for correlation analysis. The multi-objective optimization function was established. The NSGA-II. optimization algorithm was used to update the surrogate model and obtain the optimal solution, and the optimized operating parameters increased the vertical mill yield by 5.34% and the specific surface area by 9.07%. The maximum relative error between the simulated value and the optimized value is 2.02% through numerical calculation, which verifies the superiority of the optimization method of large vertical mill for performance improvement.     

基于Kriging代理模型的多点加点序列优化方法

基于Kriging 代理模型提出了一种同时考虑预测响应值及其不确定性的多点加点准则,并基于该准则发展了一套序列近似优化方法。多点加点准则基于初始样本信息和所预测的对象函数特征增加新样本集,以在寻优迭代过程中自适应地提高代理模型的精度。该文方法依据多点加点准则在一次迭代中增加多个空间无关的新样本点,适用于多机同时计算或并行计算,从而提高计算效率。以两个经典的数学函数为例,将该优化方法与期望提高准则方法进行了比较,结果表明该文提出的优化方法能够有效地提高最优解的全局性。将方法用于一盒式注塑件的成型工艺优化设计,优化结果也表明了该方法的有效性。     

含区间不确定性参数的机翼气动弹性优化

提出了一种具有区间不确定性的机翼颤振优化方法.采用拉丁超立方方法建立仿真试验表,基于MSC.Nastran平台进行颤振仿真分析.获得仿真数据之后,应用Kriging方法构造了包含区间不确定性参数的机翼颤振分析代理模型,并进行有效性检验.基于建立的代理模型并按照区间序数关系,将不确定性优化目标和约束条件转化为确定性表达形式,从面形成区间不确定性的结构优化设计方法.该方法将区间法优化和代理模型相结合,同时综合有限元仿真和遗传算法的优点,计算效率较高且应用范围较广.以某复杂机翼结构为例进行了含区间不确定性的颤振优化计算.分析结果表明了所提方法的正确性和可行性.     

基于Kriging代理模型的拉压不同模量平面问题的近似求解

该文建议采用Kriging代理模型数值求解拉压不同模量平面问题。通过本构方程光滑化、有限元法及拉丁超立方采样技术,对拉压不同模量桁架与二维平面问题,给出了基于Kriging模型的近似数值解,以代理基于有限元的数值解,并探讨了样本点数目和问题规模对所建Kriging近似模型求解精度/效率的影响。数值算例表明:所提方法可为求解拉压不同模量平面问题提供精度合理的近似数值解。当问题规模较大且正问题需要多次求解时,该方法有望显著减少计算时间,这对于降低拉压不同模量反问题与优化问题的计算开销十分重要。     

Tube hydroforming optimization using a surrogate modeling approach and Genetic Algorithm

ABSTRACT

The quality of a product obtained by hydroforming process is influenced by the geometrical, material and process parameters. In this paper, to predict an acceptable T-shaped tube with minimum wall thickness variations, and accomplishes the industrial requirements, a methodology based on the coupling of three-dimensional finite element incremental simulation based on Explicit Dynamic approach and an automatic surrogate model are proposed. The surrogate model is based on an adaptive moving target zone, and both Moving Least Square (MLS) and the Kriging technique. The optimization results are presented and compared in term of efficiency. Five quality criteria are used, an objective function defining the thickness variation with four nonlinear constraints functions, to reduce the risk of necking and to fulfil the industrial requirements. The proposed approach will provide a numerical estimation of the “best” tool dimensions and ideal punch stroke in order to obtain a final feasible workpiece.     

Multi-objective optimization of suspension parameters for rail vehicles based on a virtual prototype surrogate model

This research is intended to develop a suspension parameter optimization approach based on a virtual prototype surrogate model of rail vehicles considering the coupling effects of suspension parameters. In order to analyze the effects on the dynamic indexes, which were affected by the suspension parameters, a virtual prototype model of a rail vehicle was established. The indexes of lateral ride quality and motion stability were obtained under different combinations of suspension parameters by design of experiment and simulation of virtual prototype. For constructing objective function of multi-objective optimization model for suspension parameters, the suspension parameters that have significant effects on ride quality and motion stability simultaneously were taken as the design variables, and thereafter Kriging models of lateral ride quality index, derailment coefficient, and reduction ratio of wheel load were obtained. On this basis, the multi-objective optimization model of suspension parameters was established, in which the objective function was combined with the three Kriging models. Then, the Pareto optimal solution set and concrete value of suspension parameters were sought using the NSGA-II algorithm. The dynamic simulation results indicated that both ride quality and motion stability of the rail vehicle had been improved after the multi-objective optimization of suspension parameters.     

Crack identification based on optimization methods using harmonic elastic waves

The current paper presents a theoretical study of the identification of cracks in an elastic medium based on optimization methods using harmonic elastic waves. A generalized crack model was used to determine the elastodynamic behaviour of an arbitrarily located and oriented crack subjected to a plane harmonic wave. Sensitivity analysis was conduced to evaluate the variation of stress and strain fields with various crack parameters using the direct differential method (DDM). These general results were then used to identify an unknown crack from known strain components at discrete locations. The BFGS optimization scheme based on the sensitivity analysis was used to determine the length, the orientation and the position of the crack. Numerical simulation was conducted and the results were presented to show the effectiveness of the sensitivity analysis and the optimization method. It was observed that starting from a set of arbitrarily determined initial values, the length, position and the orientation of the crack can be accurately determined.