上海市乡村旅游景区(点)空间分布研究

引用最邻近指数、地理集中指数、基尼系数和洛伦兹曲线对上海乡村旅游景区(点)的空间分布进行测算.计算结果表明:上海乡村旅游景区(点)的最邻近指数为0.656,其空间分布呈凝聚型.其地理集中指数为87.200,这表明其空间分布是相对集聚的,并与最邻近指数的结果相映证.上海市乡村旅游景区(点)的基尼系数为0.9647,C=0.0353,这亦说明其空间分布是相对集中的,空间上表现为不均衡.且根据洛伦兹曲线,可看出,上海市乡村旅游景区(点)在空间上表现为非均衡集中分布的特征.笔者试着从相关区域内的财政收入、工业总产值、农业园区基础设施投资、常住人口密度和土地面积等因素来探讨对其空间分布结构的影响,并提出更合理的布局方案,以期促进上海乡村旅游的发展.     

随机车流作用下大跨度悬索桥振动分析

引入元胞自动机随机车流模拟方法,采用已建立的能考虑车辆空间振动的18个自由度整车模型模拟车流中重型车辆,利用单自由度车辆模型模拟车流中其他车型,模拟了能考虑邻近车辆对车流的影响桥上随机车流模型.通过车轮与桥面间的变形与接触力协调关系,建立了桥梁与车流耦合作用下运动方程.分析表明:车流中相邻车辆是否考虑对桥梁振动位移的影响较大,且随机车流作用下引起的悬索桥主梁及索塔的空间振动不能被忽视.     

加权极限学习机的多变量时间序列预测方法

提出一种基于样本分布的极限学习机预测模型WELMSD.该模型先用kN近邻密度估计方法估计出样本的密度值,再用估计出的密度值给传统ELM的经验风险项加权,克服传统ELM在对时间序列进行预测时忽略样本分布的缺点.基于Rossler混沌时间序列和上证、深证股票数据的实验仿真结果证明了所提算法的有效性,且当近邻参数k_N取值较小时,所提模型对参数不敏感,是一种更优的多变量时间序列预测模型.     

形成原子核群子结构的四项原理及其等腰三角形核素周期律

根据第四统计力学——JRG群子统计理论,首次提出了形成原子核结构的四项原理及相应的原子核周期律。这四项原理为:一是原子核有群子结构单元;二是形成原子核时群子通过热核聚合反应过程使核素结构有严格的排列顺序;三是原子核内群子结构间有动态共振作用;四是偶数群子稳定,而非偶数群子是引起总角动量和β⁺,β^-衰变的根源。基于上述四项原理,提出了核群子结构基本单元有(PB),(PB₂),(P₂B₃),并随着质子数Z的增加,核群子结构由(PB)_k过渡到(PB)_k(P₂B₃)_l;由(PB)_n(PB₂)_m过渡到(P₂B₃)_s(PB₂)_t。从而导出了k(-t)=2n(s)-Z,k(-t)=Z-2l(m)的关系式。基于此,可以画出等腰三角形原子核群子周期律。还发现,不管A,N,Z如何变化,有下列严格关系式:Z/N=(n+m)/(n+2m)或Z/N=(k+l)/(k+3l)并且Z=n+l。此公式高度地反映了所有原子核内质子和中子分布的整数规律。还可以通过等腰三角形周期律得知:k,l,m,n,s,t均与核素群子(PB),(PB₂), (P₂B₃),(P₃B₄)的2,3,5,7整数倍有关。     

非线性半参数模型最小二乘近邻估计

目前对非线性半参数模型的研究尚处于初级阶段,关于非线性半参数模型的计算理论还未见实质性报道.基于非线性半参数模型最小二乘近邻估计,给出了其参数分量和非参数分量估计的构造式,导出了参数分量和非参数分量顾及二次项直接解法的非线性直接解法.基于工程实际中的非线性模型用模拟数据进行计算,证明对于非线性半参数模型最小二乘近邻估计,可采用顾及二次项的直接解法.     

The first-principles study on the doping effect of Re in Ni3Al

Using first-principles density function for molecules method （DMol） and discrete variational method （DVM） based on the density functional theory, we studied the doping effect of Re in Ni3Al. The structure relaxation and the alloying energy show that Re has a strong Al site preference and leads to the local deformation, which is in agreement with the experimental results and other theoretical results. In addition, the charge density difference and the bond order show that Re can strongly enhance the interatomic interaction between the nearest neighbor atoms. From the density of states and the Pauli spectrum, we find that resonance states and localized states are induced by doping Re, and the doped Re atom forms the hybridized bond with the nearest neighbor atoms.     

Anomaly Detection with Artificial Immune Network

Inspired by the immune network theory, an adaptive anomaly detection paradigm based on artificial immune network, referred as APAI, is proposed. The implementation of the paradigm includes： initially, the first is to create the initial antibody network; then, through the learning of each training antigen, the antibody network is evolved and updated by the optimal antibodies. Finally, anomaly detection process is accomplished by majority vote of the k nearest neighbor antibodies in the network. The experiments used the famous Sonar Benchmark dataset in our study, which is taken from the UCI machine learning database. The obtained detection accuracy of APAI was 97.7%, which was very promising with regard to the other classification applications in the literature for this problem. In addition to its nonlinear classification properties, APAI possesses biological immune network properties such as clonal selection, immune network, and immune memory, which can be applied to pattern recognition, classification, and etc.     

最小Q-特征值第二小的给定悬挂点数的非二部单圈图

图的最小Q-特征值常被用来衡量一个图的非二部程度,受到研究者的广泛关注。在路Pn-k-2的一端接出一个圈C3,另一端接出k个悬挂边,所得的n阶图记为Ukn(3)。范益政等人最近证明Ukn(3)是最小Q-特征值达到最小的图。在他们的基础上,证明C13(n-k-1)是最小Q-特征值达到第二小的图,其中C13(n-k-1)是将Ukn(3)的一条悬挂边移至与悬挂邻点相邻的非悬挂点上所得的图。     

基于距离的相似最近邻搜索算法研究

为了提高相似最近邻搜索（ANN）算法的精度,提出了一种在度量空间下基于距离的相似最近邻搜索算法-优化的VP森林（OVF）算法。在传统VP树（VT）算法的基础上,首先采用改进的选择优势点的方法,通过从数据集采样优势点候选集,对其进行评估,选取其中区分度大的点作为优势点;然后提出构建多棵VP树的新方法,改进距离优势点远的子树中最近邻不紧凑问题;接着提出使用优先队列与剪枝搜索方法结合的新搜索方法查找最近邻,减少了很多不必要的距离计算。最后通过实验结果表明,本文方法在数据维度、数据集大小、返回不同邻居个数、不同的距离函数及建树个数方面精度有了很大的提高。     

非晶和纳米ZrO2.Y2O3的EXAFS研究

利用扩展X射线吸收精细结构技术对用化学共沉淀法制备的非晶和纳米ZrO2.15%Y2O3体系进行了研究，分析显示，在从非晶态向纳米结构晶化的结构中，最后邻的Zr-O配位层的配位数和键长没有发生明显的改变，说明300度时已经形成和900度相同的最近邻局域结构，而对于Zr-Zr(Y)配位层，随着晶粒尺寸的减少，配位数明显降低，键长显著缩短，具有比Z r-O配位层更大的无序，这些结果表明晶粒尺寸的变化对于较远配位层的影响要比对最近孤影响大得多。     

Classification by ALH-Fast Algorithm

The adaptive local hyperplane (ALH) algorithm is a very recently proposed classifier, which has been shown to perform better than many other benchmarking classifiers including support vector machine (SVM), K-nearest neighbor (KNN), linear discriminant analysis (LDA), and K-local hyperplane distance nearest neighbor (HKNN) algorithms. Although the ALH algorithm is well formulated and despite the fact that it performs well in practice, its scalability over a very large data set is limited due to the online distance computations associated with all training instances. In this paper, a novel algorithm, called ALH-Fast and obtained by combining the classification tree algorithm and the ALH, is proposed to reduce the computational load of the ALH algorithm. The experiment results on two large data sets show that the ALH-Fast algorithm is both much faster and more accurate than the ALH algorithm.     

Linear and non-linear pseudocapacitances with or without diffusion control

Pseudocapacitance is an important reversible charge storage mechanism in many electrode materials. Although the concept was first proposed in early 1960s, it has been more widely studied following the observation of rectangular cyclic voltammograms (CVs) when testing some transition metal oxides and electronically conducting polymers, and the association with supercapacitor. However, interpretation of pseudocapacitance is inconsistent in the literature. Although all agree that materials are pseudocapacitive if they undergo Faradaic reactions and exhibit rectangular CVs, some have regarded any surface confined Faradaic reactions which may present non-rectangular or even peak-shaped CVs to be pseudocapacitive. In the case of rectangular CVs, the amount of charge stored in the electrode is a linear function of the electrode potential, whilst for non-rectangular or peak-shaped CVs, the relationship is non-linear. It is shown in this article that only linear pseudocapacitance is of relevance to supercapacitor, but non-linear pseudocapacitance may find applications in rechargeable battery and supercapattery. Further, it is clarified that the equation i ​= ​k₁v ​+ ​k₂v^1/2 is useful in analysis of electrode kinetics in terms of surface confinement and diffusion control. However, this kinetic equation is blind to the thermodynamically determined charge storage mechanisms as shown by experimental evidence, and should not be used to differentiate non-capacitive Faradaic processes from pseudocapacitance, either linear or non-linear.     

基于动态加权差分主成分分析的故障检测

针对工业过程中存在的动态特性和多模态特性问题，提出一种动态加权差分主成分分析法 (dynamic weighted differential principal component analysis, DWDPCA)。首先通过设置合理的时间窗描述系统的时序特性；其次对时间窗内的样本寻找第一近邻和第一近邻的近邻集，使用加权差分法对数据进行处理，解决数据中心漂移问题；最后利用处理好的数据建立主成分分析（principal component analysis，PCA）模型进行故障检测。该方法可解决数据动态、中心漂移问题。使用该方法对数值例子和田纳西-伊斯曼TE(tennessee eastman)过程进行故障检测验证所提出方法的有效性。     

计时电量法测定4,5-二氮芴-9-酮的扩散系数和反应速率常数

运用循环伏安法测定4,5-二氮芴-9-酮（DAFO）在DMF溶液中不同扫描速率时的循环伏安图,出现2对峰形良好的氧化还原峰;采用计时电量法通过控制阶跃电位,测得不同温度下DAFO在DMF溶液中的扩散系数和反应速率常数,293K时扩散系数为5.39×10-6cm2·s-1;当选择阶跃电位使电极过程处于扩散和电化学混合控制时,测得不同电位的反应速率常数k_f. 根据不同温度下的标准速率常数k₀,得到表观活化能E_a为10.5kJ·mol-1.      

Super-Edge-Connectivity of G(k,d,s)(s≥k/2)

A graph G is super-edge-connected,for short super-λ,if every minimum edge-cut consists of edges adjacent to a vertex of minimum degree.Alphabet overlap graph G(k,d,s)is undirected,simple graph with vertex set V={v|v=1()kv…v;vi∈{1,2,…,d},i=1,…,k}.Two vertices u=(u1…uk)and v=(v1…vk)are adjacent if and only if us+i=vi or vs+i=ui(i=1,…,k-s).In particular G(k,d,1)is just an undirected de Bruijn graph.In this paper,we show that the diameter of G(k,d,s)is k s,the girth is 3.Finally,we prove that G(k,d,s)(s≥k/2)is super-λ.     

基于最优投影和动态阈值的最近邻搜索算法

作者在前人工作成果的基础上,提出并实现了一种基于最优投影和动态阈值调整的最近邻搜索算法DTA(Dynamic Threshold Algorithm);证明了最优投影线定理和投影邻域定理;并分析了DTA算法与SNN算法相比在算法性能上的优势.实验结果表明,当数据规模增大时,DTA算法的运行时间增加相对缓慢,在大规模数据集上DTA算法的运行时间可达传统算法的10%以下;DTA算法对阈值的变化不敏感,能适应不同分布的数据集合.     

金刚石单空位的分子动力学模拟

用分子动力学方法模拟了金刚石中单空位的最近邻原子和次近邻原子的弛豫过程,计算所用的碳原子间相互作用势为Tersoff多体经验势.结果表明,单空位的最近邻原子是向外弛豫的,平均弛豫幅度约为2×10     

甲醇与O[3P]反应机理的从头算及动力学

采用UQCISD/ 6-311G (d,p )从头算方法,优化甲醇和O [³P ]的反应两个通道、反应物、过渡态和产物的几何构型。进一步运用G2方法进行单点能量校正,得出通道 (1)和通道 (2)的位垒分别是48.86kJ/mol和28.89kJ/mol。并指出通道 (1 )是吸热反应,而通道 (2 )是放热反应。在300～3200K温度范围内,采用传统过渡态理论计算两个反应通道各自的速率常数k₁ 和k₂ ,由此采用非线性最小二乘法,得出这两个反应通道各自的速率方程为k₁=2.43×10^-18×T^2.23×exp(- 32.97/T)cm³mol^-1 s^-1 (300K≤T≦3200K), k₂=6.12× 1 0 ^-18×T^2.19×exp(- 1396/T)cm³mol^-1s^-1(300K≤T≦3200K) 2/k₁对温度变化的依赖关系。计算得出CH₃OH和O[³P]反应的总速率常数k₁₊₂ ,与实验结果取得很好的一致。     

基于分块2D DCT和(2D)~2PCA的人脸识别

人脸识别是生物特征识别技术中的重要研究领域,应用前景广阔.研究者们虽然提出了很多人脸识别算法,但其性能仍需进一步改进.为了提高现有人脸识别算法的识别准确率,提出了一种新的基于分块二维离散余弦变换(2DDCT)和双向二维主成分分析((2D)2PCA)的人脸识别算法.首先,将图像分块,利用2DDCT进行图像压缩,去除冗余信息,并通过逆2DDCT重建图像;其次,通过(2D)2PCA消除图像的行、列相关性,降低特征维数;最后,应用最近邻分类器进行人脸识别,在ORL人脸数据库中的实验证明了本算法的有效性.     

用周期模型和近邻算法预测话务量时间序列

客服中心话务量虽然具有周期性,但在不同时间遵循不同变化规律,这是话务量预测的难点。针对这个问题,以某电信公司一年的实际话务数据为基础,分别采用周期模型和基于实例的近邻算法进行话务量时间序列预测,并对比分析了两种预测方法的效果。实验数据表明,对工作日话务量的预测,周期模型的预测效果优于近邻算法;对非工作日话务量的预测,近邻算法的预测效果优于周期模型。为取得更好的预测效果,实现了周期模型和近邻算法相结合的预测方法。结果表明,在最好的情况下,该方法的预测精度比周期模型提高约19.7%,比近邻算法提高约48.8%。