基于最小二乘法的灰色组合预测模型及其应用

根据GM(1,1)模型原理,构建了灰色增量模型和灰色组合预测模型.分别利用中国和河南1986-2015年的人口数据建立模型,采用最小二乘法求解灰色组合预测模型的最优权系数.通过GM(1,1)模型、灰色增量模型和灰色组合预测模型对上述算例进行误差分析,实验结果表明灰色组合预测模型预测精度明显的优于其它单项预测模型.     

一种新的组合模型在信用评分中的应用

针对信用评分的特点,首先对传统的Bagging算法进行改进,然后基于不同的基分类器构建多个改进的Bagging组合模型,最后根据一种全新的方式将其组合,得到双层组合模型.通过实证分析得出以下结论:一、与原Bagging算法相比,改进算法能有效降低第二类错误率;二、与单层组合模型相比,多层组合模型能进一步降低第二类错误率和损失率.     

含MCVaR的投资组合优化统一模型

研究了Duarte提出的投资组合优化统一模型及条件风险价值(CVaR),分析了以CVaR为风险度量的投资组合优化模型的具体形式,建立了统一七种模型的投资组合优化统一模型,并发现统一模型是一个凸二次规划问题.     

基于等效原理的波形钢腹板组合箱梁桥动力特性分析

为了简化波形钢腹板组合箱梁的三维有限元建模过程,通过刚度和位移等效的方法将三维波形钢腹板有限元模型简化为二维正交各向异性板的有限元模型.简化之后的有限元模型几何外形简单,单元和自由度的数量大幅降低,减少了计算时间,提高了计算效率.通过对比三维波形钢腹板组合箱梁有限元模型的自振频率计算结果、二维等效正交各向异性板组合箱梁有限元模型的自振频率计算结果以及实测频率值,发现三者吻合良好,验证了等效模型有限元建模方法的正确性和可靠性.研究结论可为波形钢腹板组合箱梁桥提供一种简单的有限元建模方法.     

最佳灰色回归组合模型及其在中国火灾预测中的应用

火灾每年给国家和人民生命财产造成巨大损失.火灾现象具有随机性、模糊性,是个复杂的灰色系统行为.研究火灾发生规律及发展趋势,具有实用价值.为此,首先给出最小二乘估计(LSE)意义下的最佳组合预测模型的定义,并求得组合模型的权的公式和证明权的唯一性.其次,用回归分析方法建立多个回归模型,并按以下三条标准:①回归指数(或相关系数)r大、②系统误差s小、③模型精度p高,选定最佳非线性回归模型;用灰色理论建立多个灰色模型,并按以下三条标准:①后验差比值c小、②小误差概率P大、③预测关联度ξ大,选定最佳灰色模型;再用最小二乘法将最佳回归模型与最佳灰色模型有机地结合起来建立的中国火灾最佳灰色回归组合预测模型.最佳灰色回归组合预测模型综合利用前两者提供的不同的有用信息,改善了单一模型的局限性,提高了模型的预测精度,减少了预测误差,使预测效果更佳.组合模型预测中国年火灾起数处于动态增长过程.     

均值-方差-近似偏度投资组合模型和实证分析

均值-方差投资组合模型作为现代投资组合理论的基础, 采用方差作为风险度量,但忽略了投资组合收益的非对称性. 而考虑收益非对称性的基于偏度的投资组合模型由于非凸和非二次性 使模型难以求解. 本文提出用上下半方差的比值近似刻画偏度, 建立了均值-方差-近似偏度(MVAS）模型,并利用该模型对中国证券市场主要股票指数进行实证分析. 实证分析结果表明, 在收益率非正态分布的市场中,考虑了收益率非对称性的投资组合模型较传统的MV和MAD模型具有更优的表现.     

基于LS-SVM的管道腐蚀速率灰色组合预测模型

为提高管道腐蚀速率预测精度,建立了一种基于最小二乘支持向量机的灰色组合预测模型.以各种灰色模型对管道腐蚀速率的预测结果作为支持向量机的输入,以管道腐蚀速率的实测值作为支持向量机的输出,采用最小二乘支持向量机回归算法和高斯核函数对支持向量机进行训练,利用训练好的支持向量机进行组合预测.预测模型兼具灰色模型所需原始数据少、建模简单、运算方便的优势和最小二乘支持向量机具有泛化能力强、非线性拟合性好、小样本等特性,弥补了单一预测模型的不足,避免了神经网络组合预测易于陷入局部最优的弱点.模型结构简单、实用,仿真结果验证了其有效性.     

基于初值修正的组合灰色Verhulst模型

对灰色Verhulst模型的拟合精度进行了分析,表明初值的选取对模型精度有重要影响.为提高灰色模型的拟合精度,利用最小二乘原理确定初值中的待定常数,给出了初值修正的灰色Verhulst模型.进一步利用修正模型和传统灰色Verhulst模型建立了组合灰色Verhulst模型,在平均相对误差最小的原则下,利用蚁群算法确定组合权系数.最后通过两个应用实例进行了计算和分析,结果表明,通过初值修正和组合模型能够提高灰色Verhulst模型的拟合精度,便于通过程序实现.     

摩擦市场的均值-离差证券组合选择模型

本文给出了基于历史收益率数据的均值 -平均绝对离差型证券组合投资模型 .该模型采用收益的平均绝对离差作为风险的尺度 ,可以通过求解线性规划来获的摩擦市场 (如具有税收和交易费 )最优投资组合 ,避免了均值 -方差模型求解二次规划的复杂性 .     

具有投资限制的风险资产模糊随机投资组合模型

研究了模糊随机环境下风险资产投资组合选择问题.利用模糊随机变量刻画风险资产的收益率,建立了具有投资限制的风险资产投资组合选择的一般模糊随机均值-方差模型,该模型包括了是否允许卖空及具有投资比例下界约束的情况.在此基础上,提出了具有梯形模糊随机收益率的具体投资组合优化模型,这些模型能够转化为二次规划问题求解.最后,利用上证50指数中的9种股票对模型进行了实证分析,结果表明模型能够有效分散非系统性风险.     

Partial hybrid finite elements for composite laminates

Three types of partial hybrid finite elements are presented in order to set up a global/local finite element model for analysis of composite laminates. In the global/local model, a composite laminate is divided into three different regions: global, local, and transition regions. These are modeled using three different elements. In the global region, a 4-node degenerated plate/shell element is used to model the overall response of the composite laminate. In the local region, a multilayer element is used to predict detailed stress distribution. In the transition region, a multilayer transition element is used to smoothly connect the two previous elements. The global/local finite element model satisfies the compatibility of displacement at the boundary between the global region and the local region. It also satisfies the continuity of transverse stresses at interlaminar surfaces and traction conditions on the top and bottom surfaces of composite laminates. The global/local finite element model has high accuracy and efficiency for stress analysis of composite laminates. A numerical example of analysis of a laminated strip with free edge is presented to illustrate the accuracy and efficiency of the model.     

A finite element model based on statistical two-scale analysis for equivalent heat transfer parameters of composite material with random grains

In this paper, a new finite element model based on statistical two-scale analysis for predicting the equivalent heat transfer parameters of the composite material with random grains is presented and its convergence, its error result and the symmetry, positive property of equivalent heat transfer parameters matrix are also proved. Firstly, some definitions of the probability space and the composite material with random grains are described and the STSA formulation predicting the equivalent heat transfer parameters of the composite material are briefly reviewed. Next, a finite element formulation and its corresponding procedure for the composite material with random grains is described. Then, the convergence, the error estimate and the symmetry, positive property of the equivalent heat transfer parameters matrix computed by FE based on STSA are proved. The numerical result shows the validity of the FE model based on STSA and the convergence and the symmetry, positive property of the equivalent heat transfer parameter matrix of the composite material with random grains by the FE model.     

基于随机电阻网络碳毡复合层力阻建模

基于碳纤维单丝力阻效应影响因素的统计规律,揭示了碳毡复合层力阻传感特性的导电网络定量化模型．对碳纤维单丝相互搭接后的长度进行了讨论,得到关于纤维均长、纤维密度、纤维根数等多参量间的统计关系;在假设毡内纤维随机均匀分布且搭接点正态分布的情况下,分别对4种不同尺寸的碳纤维毡复合层的电阻进行了估测,并通过箱线图分析表明实测值在相应的估测范围内,说明本模型可较好地对碳纤维随机乱相分布的碳毡复合层电阻进行估测;最后,基于碳毡电阻网络的统计模型,对碳毡复合层的力阻特性进行了研究,通过比较理论电阻变化率与实测电阻变化率发现模型给出的复合层力阻效应灵敏系数较为稳定,从而表明随机电阻网络模型可较好的模拟碳毡受载荷时的力阻行为．     

Finite-element modeling of the particle clustering effect in a powder-metallurgy-processed ceramic-particle-reinforced metal matrix composite on its mechanical properties

A new numerical method is proposed to predict the effect of particle clustering on grain boundaries in a ceramic- particle-reinforced metal matrix composite on its mechanical properties, and micromechanical finite-element simulation of stress–strain responses in composites with random and clustered arrangements of ceramic particles are carried out. A particular material modeled and analyzed is a TiC-particle-reinforced Al matrix composite processed by powder metallurgy. A representative volume element of a composite microstructure with 5 vol.% TiC is reconstructed based on the tetrakaidecahedral grain boundary structure by using a modified random sequential adsorption. The model proposed in this study accurately represents the stress concentrations and particle-particle interactions during deformation of the powder-metallurgy-processed composite. A comparison with the random-arrangement model shows that the present numerical approach is more accurate in simulating the behavior of the composite material.     

The suboptimality of composite forecasts derived from posterior probabilities

There has been quite a common practice of formulating composite forecasts using the posterior probabilities derived from Bayesian model discrimination methods. It is shown that such composite forecasts are normally suboptimal in the sense that there exists an alternative composite forecast with a strictly lower expected error.     

A new look at forecasting annual corporate earnings in the U.S.A.

In this study, composite earnings per share models are estimated for 35 chemical, food, and utility firms during the 1981–1982 period. Although it is generally held that financial analysts produce superior earnings forecasts when compared to time series model forecasts, the results of this study indicate that analysts fared very poorly in 1982 and the average mean square forecasting error of analyst forecasts may be reduced by 74.2 percent by combining analyst and univariate time series model forecasts. This reduction is very interesting when one finds that the univariate time series model forecasts do not substantially deviate from those produced by random walk drift models, the ARIMA (0, 1, 1) process. Moreover, despite the high degree of correlation existing among analyst and time series forecasts, the ordinary least squares estimation of the composite earnings model is a better forecasting model than the composite earnings models estimated with ridge regression and latent root regression techniques.     

Deep drawing of a layered composite: material characterization and finite-element simulation

Composite materials are widely used in different industrial fields, because of their good formability and their high strength to weight ratio. In the present work a triple-layered sandwich composite is investigated. Experimental tests at room temperature are carried out for the materials constituting the composite. A finite element model of a deep-drawing process of the composite is performed, where a finite strain constitutive model for the metal part, with material parameters calibrated to uniaxial tensile tests, has been implemented. Experimental results are compared to the numerical simulations in view of validation purposes. (© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim)     

A micromechanics model for FRC composites

A multiscale model for FRC composite structures taking into consideration the complex interactions at the scales of the fiber and microcracks is proposed. At the scale of the single fiber, a semi-analytical model characterizes the microslip behavior at the interface between the matrix and the fiber in terms of the overall composite stresses. The influence of fiber bundles on microcrack bridging and arrest is taken into account within the framework of linear elastic fracture mechanics. Upscaling to the macroscopic level using continuum micromechanics shows that the macroscopic deformation of the FRC composite is governed by a ’TERZAGHI’ like effective stress. For the finite element analyses of failure behavior at the scale of the composite structure, an ’interface solid element’ technique is used to consider localized cracking. Selected numerical and semi-analytical results together with experimental validations are provided. (© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim)