应力应变顺序对应规律及其学术意义

应力应变顺序对应规律,是王仲仁教授于1979年提出的,起初在国内有一些争议,后来王仲仁教授又作了理论推导与实验验证。目前,该规律已被列入《金属塑性成形原理》全国通用教材。本文阐述此规律的提出和完善过程,并讨论其学术意义。     

Effects of γ/γ interfacial structures on continuous coarsening of lamellar microstructure in TiAl alloy during aging

The effects of γ/γ interfacial structures on continuous coarsening of the fully lamellar microstructure in Ti-48Al alloy aged at 1 150 ℃ were investigated by using transmission electron microscopy(TEM). Continuous lamellar coarsening can be achieved not only by migration of interface faults(such as ledges, edges and curved interfaces) but also by migration and decomposition of perfect γ/γ lamellar interfaces. Thermal grooves, initiative positions of interfacial dissociation, can frequently form at the triple point junctions between the 120°-rotational ordered γ domain boundaries within γ lamellae and the lamellar interfaces. During the early stage of aging at 1 150 ℃, the interface migration and dissociation took place preferentially at the 120°-rotational ordered lamellar interfaces. Comparing the relative thermal stability of the true-twin, pseudo-twin and 120°-rotational ordered γ/γ lamellar interfaces shows that the 120°-rotational ordered lamellar interface is the most unstable. The reason of this phenomenon was analyzed through the comparisons of the interfacial energies and atomic arrangements of the three types of γ/γ lamellar interfaces.     

AB_5型贮氢合金P－C关系的人工神经网络研究

本文利用人工神经网络方法研究了ＡＢ５型贮氢合金的Ｐ－Ｃ曲线、平台宽度和平台压力等，并取得了满意的结果，表明人工神经网络可用于贮氢材料的设计     

基于支持向量机(SVM)的祁连山典型小流域日降水-径流模拟研究

及时准确的日径流预测在流域水资源的合理规划、利用及管理中具有十分重要的作用。本文以支持向量机(SVM)模型为基础,以祁连山典型小流域-排露沟流域为研究区域,建立了流域日降水-径流模型,对流域未来1~7 d的日径流量进行了模拟预测。为检验SVM模型的有效性,模拟结果与人工神经网络(ANN)模型预测结果进行了对比。结果表明:SVM和ANN均表现出了很高的精度;但相比于传统的ANN模型,SVM模型的预测精度显著提高。表明SVM模型在半干旱山区小流域径流预测中有更好的适用性,可以用于流域中长期日径流预测,是资料有限的条件下中长期日径流预测的有效工具。     

Automatic generation algorithm of expected results for testing of component-based software system

ContextA component-based software (CBS) system is a typical example of a composite component, which is composed of two or more atomic components. In the test of a CBS system, it is necessary to automatically generate expected results because they are compared with the actual results.ObjectiveThis paper proposes an automatic generation algorithm of expected results for the testing of a CBS system.MethodThe algorithm analyzes Input/Output (I/O) relationships of a CBS system to identify inputs that influence its outputs. Then, the algorithm uses test cases of atomic components for each input and automatically generates expected results. To evaluate the proposed approach, we compare the proposed approach with the other I/O relationship based approach via two case studies which are the CBS systems for guide robot. The comparison shows the effectiveness of our approach.ResultsTo verify effectiveness of the proposed approach, we measure the cost of manual generation of expected results for each case study. The costs are from 0.001 to 0.015 in the first case study, and are from 0.590 to 0.998 in the second case study.ConclusionThis proposed approach reduces the time-consuming and error-prone task which manually generates expected results for the testing of a CBS system.     

基于半监督图聚类的项目主题模型构建方法

项目文档主题表征的好坏直接影响后续评审专家的推荐效果.为有效利用项目文档片段之间的关联关系进行项目主题分析,提出一种基于半监督图聚类的项目主题模型构建方法.该方法首先分析项目文档的结构特点,提取项目名称、项目关键字等能表征主题的结构信息,结合专家证据文档、专家主题关系网等能表征专家主题的外部资源,定义及提取项目文档片段之间的关联关系特征;然后,利用不同类型的关联关系计算项目文档片段之间的相关性,构建项目文档片段间的无向图模型;最后,利用已标记关联关系特征作为聚类的监督信息,采用半监督图聚类算法对项目文档片段进行聚类,从而实现项目主题的提取.项目主题提取对比实验结果验证了所提方法的有效性,项目文档结构化特征、专家证据文档以及专家主题关系网对项目主题模型的构建具有一定的指导作用.     

一种基于GPS轨迹的道路拓扑生成方法

复杂路网拓扑的自动生成建立在道路提取和交叉路口识别的基础之上,是智能交通控制和自动导航服务等领域的研究热点之一,基于浮动车或出租车的GPS轨迹可以反映交通路网的拓扑结构。为此,提出了一种基于GPS轨迹的道路拓扑生成方法,即在无道路地图辅助的情况下,该方法基于大规模GPS轨迹,能够快速提取路口交叉点,自动构建具有地理位置信息的拓扑结构和计算相邻路口的网络距离。实验结果表明,该方法能够提取出各个道路交叉点并建立各点之间的拓扑关系。在提取主干道路拓扑实验中,在设置路宽为55米的情况下提取路口交叉点的正确率达到了87.08%,各路口之间的平均网络距离误差率为8.87%,并且能够正确地得到交叉点之间的连通关系。     

一种基于局部加权回归的分类方法

分类是数据挖掘和数据分析中最有应用价值的技术之一。传统的积极学习方法需要预先对模型空间进行假设,并且没有充分考虑到实例之间的相关性,其泛化能力将会受到一定程度的影响。针对上述问题,提出了一种基于新型映射关系的局部加权回归方法 MLWR。该算法首先找出测试样本在训练集中的近邻样本,然后建立测试样本和近邻样本的回归函数,根据建立的回归模型和近邻样本的标签,计算得到测试样本的标签。实验与当前流行的多种分类方法在UCI的9个数据集上进行测试。实验结果表明我们的方法能有效地提高分类精度,对较大样本数据也有较好的适用性。     

基于3D-DCT和SVD的鲁棒彩色图像水印算法

在数字图像水印领域,水印算法主要集中于灰度图像,且提出的大部分彩色图像水印算法往往仅在亮度分量或在彩色图像的每一通道中嵌入水印,未能充分利用彩色图像的冗余空间,影响了水印的透明性和鲁棒性。针对此问题,提出了一种新颖的基于三维离散余弦变换和奇异值分解的彩色图像水印算法。算法先对水印图像进行预处理和对彩色图像进行互不重叠的分块;其次对每一分块进行三维离散余弦变换;最后选择对三维离散余弦变换系数的第一分量进行奇异值分解。嵌入水印时,对三维离散余弦变换系数第一分量的最大奇异值和第二分量分别采用量化和关系的嵌入方法嵌入水印。提取水印时,分别采用量化和关系提取算法提取水印并进行比较,选取相似值高的水印图像作为最终提取的水印。实验结果表明,提出的算法具有较好透明性的同时,具有抵抗常规信号处理和模糊、扭曲及锐化等攻击的能力。     

Neural Network and Classification Approach in Identifying Customer Behavior in the Banking Sector: A Case Study of an International Bank

The customer relationship focus for banks is in development of main competencies and strategies of building strong profitable customer relationships through considering and managing the customer impression, influence on the culture of the bank, satisfactory treatment, and assessment of valued relationship building. Artificial neural networks (ANNs) are used after data segmentation and classification, where the designed model register records into two class sets, that is, the training and testing sets. ANN predicts new customer behavior from previously observed customer behavior after executing the process of learning from existing data. This article proposes an ANN model, which is developed using a six‐step procedure. The back‐propagation algorithm is used to train the ANN by adjusting its weights to minimize the difference between the current ANN output and the desired output. An evaluation process is conducted to determine whether the ANN has learned how to perform. The training process is halted periodically, and its performance is tested until an acceptable result is obtained. The principles underlying detection software are grounded in classical statistical decision theory.