用相指示器函数方法(分散元概率密度函数)推导及验证湍流稀两相流动的欧拉-欧拉模型

A statistical formalism overcoming some conceptual and practical difficulties arising in existing two-phase flow (2PHF) mathematical modelling has been applied to propose a model for dilute 2PHF turbulent flows.Phase interaction terms with a clear physical meaning enter the equations and the formalism provides some guidelines for the avoidance of closure assumptions or the rational approximation of these terms. Continuous phase averaged continuity, momentum, turbulent kinetic energy and turbulence dissipation rate equations have been rigorously and systematically obtained in a single step. These equations display a structure similar to that for single-phase flows.It is also assumed that dispersed phase dynamics is well described by a probability density function (pdf) equation and Eulerian continuity, momentum and fluctuating kinetic energy equations for the dispersed phase are deduced.An extension of the standard k-c turbulence model for the continuous phase is used. A gradient transport model is adopted for the dispersed phase fluctuating fluxes of momentum and kinetic energy at the non-colliding, large inertia limit. This model is then used to predict the behaviour of three axisymmetric turbulent jets of air laden with solid particles varying in size and concentration. Qualitative and quantitative numerical predictions compare reasonably well with the three different sets of experimental results, studying the influence of particle size, loading ratio and flow confinement velocity.     

A hybrid SVM and supervised learning approach to fuzzy min‐max hyperbox classifiers

Abstract

In this paper, a fuzzy min‐max hyperbox classifier is designed to solve M‐class classification problems using a hybrid SVM and supervised learning approach. In order to solve a classification problem, a set of training patterns is gathered from a considered classification problem. However, the training set may include several noisy patterns. In order to delete the noisy patterns from the training set, the support vector machine is applied to find the noisy patterns so that the remaining training patterns can describe the behavior of the considered classification system well. Subsequently, a supervised learning method is proposed to generate fuzzy min‐max hyperboxes for the remaining training patterns so that the generated fuzzy min‐max hyperbox classifier has good generalization performance. Finally, the Iris data set is considered to demonstrate the good performance of the proposed approach for solving this classification problem.     

ModEnPBT: A Modified Backtracking Ensemble Pruning algorithm

This paper proposes a new Modified Backtracking Ensemble Pruning algorithm (ModEnPBT), which is based upon the design idea of our previously proposed Ensemble Pruning via Backtracking algorithm (EnPBT), and however, aiming at overcoming its drawback of redundant solution space definition. Solution space of ModEnPBT is compact with no repeated solution vectors, therefore it possesses relatively higher searching efficiency compared with EnPBT algorithm. ModEnPBT still belongs to the category of Backtracking algorithm, which can systematically search for the solutions of a problem in a manner of depth-first, which is suitable for solving all those large-scale combinatorial optimization problems. Experimental results on three benchmark classification tasks demonstrate the validity and effectiveness of the proposed ModEnPBT.     

SCX超细分级机气流预分散特性与数值模拟研究

运用CFD软件对预分散装置的几何结构进行数值模拟,建立了局部颗粒浓度与平均颗粒浓度的的关联曲线来表征分散性的好坏,得出了一组较优设计目标。以重质碳酸钙为物料在SCX400超细分级机上进行筛余实验并绘制Tromp曲线,通过对实验结果的分析定性验证了数值模拟的正确性。     

On the statistical physics modeling of dye adsorption onto anionized nylon: Consequent new interpretations

Experimental adsorption isotherm of two basic dyes: Basic Blue 3 and Basic Red 24 from aqueous solution onto modified nylon 6,6 were analyzed by using a double layer adsorption model with two energy levels. Such model is based on statistical physics treatment. The parameters involved in the analytical expression of this model such as the fraction or the number of adsorbed dye molecule(s) per site, n, the receptor sites density, N_M, and the energetic parameters, c₁ and c₂, were determined by fitting the experimental adsorption isotherms at four temperatures between 293 and 353 K with different degrees of grafting between 20 and 80%. The evolution of these parameters versus temperature and the grafting percent allows us to interpret and better understand this adsorption process at molecular level. Two different behaviors of the two dye molecules were highlighted according to their localized and non localized charges. The configurational entropy at various temperatures has also been studied. This parameter allowed to deduce some results related to the evolution of the disorder at the adsorption surface. © 2011 Wiley Periodicals, Inc. J Appl Polym Sci, 2012     

On output independence and complementariness in rank-based multiple classifier decision systems

This study presents a theoretical analysis of output independence and complementariness between classifiers in a rank-based multiple classifier decision system in the context of the partitioned observation space theory. To enable such an analysis, an information theoretic interpretation of a rank-based multiple classifier system is developed and basic concepts from information theory are applied to develop measures for output independence and complementariness. It is shown that output independence of classifiers is not a requirement for achieving complementariness between these classifiers. Namely, output independence does not imply a performance improvement by combining multiple classifiers. A condition called dominance is shown to be important instead. The information theoretic measures proposed for output independence and complementariness are justified by simulated examples.     

基于韦伯梯度方向直方图的人脸识别算法

针对传统人脸识别算法在姿态、表情和光照等变化下而引起识别效果不佳的问题,提出一种韦伯梯度方向直方图人脸识别算法（HWOG）。利用差动激励提取图像的结构和纹理信息,利用HOG算子提取原始图像的边缘特征,分块统计直方图特征信息,将所有分块的直方图串接得到人脸图像HWOG特征,用最近邻分类器进行分类。在YALE人脸库、ORL人脸库上和CAS-PEAL-R1进行实验,实验结果表明所提算法能有效提高识别率,且对光照、表情和姿态变化有较好的鲁棒性。     

局部方向梯度幅值与相位差分的人脸识别算法

针对传统人脸识别算法在单训练样本下效果不佳,提出一种局部方向梯度幅值和相位差分相结合的方法（LDGMPD）,首先提取图像的梯度幅值与相位,梯度幅值图像与8个Kirsch模板卷积得到每个子邻域的8个边缘梯度值;然后对相位进行局部差分。局部方向梯度幅值与相位差分仅使用边缘梯度值与相位局部差分值中最大值的方向编码成一个二位八进制数,产生LDGMPD值。再选取结构对比信息对各LDGMPD人脸分块进行加权处理,提取人脸的LDGMPD直方图特征,最后利用最近邻分类器分类识别。在AR和CAS-PEAL-R1共享库上进行实验表明LDGMPD在单样本人脸识别具有较好的效果。     

基于集成学习的多重集典型相关分析方法

特征提取是模式识别中的关键问题之一,对提高系统分类性能具有重要意义。常用的特征提取方法包括主成分分析、线性鉴别分析、典型相关分析等等,多重集典型相关分析是基于传统的典型相关分析基础上发展而来,利用多组（大于2）特征数据集进行特征提取。基于集成学习的多重集典型相关分析的方法（EMCCA）,是通过将样本化分成若干小的样本,形成若干个特征数据集,利用多重集典型相关分析对这组数据集做特征提取,并结合集成学习对样本进行分类。在UCI上的多特征手写体数据集上的实验结果表明：相比于传统的PCA,CCA特征提取方法,多重集典型相关分析具有更优的特征提取效果,结合集成学习后具有更好的分类效果。