Hierarchical nanosheet-based CoMoO4–NiMoO4

Hierarchical nanosheet-based CoMoO4–NiMoO4
Hierarchical nanosheet-based CoMoO4–NiMoO4

分层线性模型

分层线性模型(hierarchical linear model HLM)的原理及应用 一、概念: 分层线性模型(hierarchical linear model HLM)又名多层线性模型(Multilevel Linear Model MLM)、层次线性模型(Hierarch Linear Mode1)、多层分析(Multilevel Analysis/Model)。相对于传统的两种统计方法:一般线性模型(general linear model GLM)和广义线性模型(generalized linear models GLMs),它们又有所不同,HLM中的线性模型指的是线性回归,不过它与一般的分层线性回归(Hierarchical Regression)又是不同的,具体的不同见下面数学模型部分。HLM又被通俗的称为“回归的回归”。 Wikipedia:“一般线性回归和多重线性回归都是发生在单一层面,HLM相对于更适用于嵌套数据(nest data)。” 在理解HLM之前应了解有关回归分析和嵌套设计(分层设计)的基本知识。 二、模型: 1、假设:由于个体行为不仅受个体自身特征的影响,也受到其所处环境(群体/层次)的影响。相对于不同层次的数据,传统的线性模型在进行变异分解时,对群组效应分离不出,而增大模型的误差项。而且不同群体的变异来源也可能分布不同,可能满足不了传统回归的方差齐性假设。在模型应用方面,不同群体(层次)的数据,也不能应用同一模型。鉴于传统方法的局限性,分层技术则解决了这些生态谬误(Ecological Fallacy)。它包含了两个层面的假设: a、个体层面:这个与普通的回归分析相同,只考虑自变量X对因变量Y的影响。 b、群组层面:群组因素W分别对个体层面中回归系数和截距的影响。 2、数学模型: a、个体层面: Yij=Β0j+Β1jXij+eij b、群组层面: Β0j=γ00+γ01Wj+U0j Β1j=γ10+γ11Wj+U1j 涉及到多个群组层次的时候原理与之类似,可以把较低级层次的群组,如不同的乡镇层面与不同的县市层面,可以这样理解,乡镇即是一个个体,群组即是不同的县市。更多层次的可以这样理解,一直是下一层对上一层回归系数和截距的回归。与普通的“回归的回归”不同的是,整个计算过程通过迭代过程完成。 3、因变量: 此处数学模型仅适用于连续的单因变量。非连续因变量、多因变量、潜变量以及非典型的嵌套设计,多层线性模型也可以进行处理,但对模型的设定会更复杂。 4、与分层回归的区别: a、向前回归、向后回归和逐步回归: 向前回归:根据自变量对因变量的贡献率,首先选择一个贡献率最大的自变量进入,一次只加入一个进入模型。然后,再选择另一个最好的加入模型,直至选择所有符合标准者全部进入回归。

Wincc V7.3 采集excel文件数据填入Microsoft Hierarchical Flexgrid

Wincc V7.3 采集excel文件数据填入Microsoft Hierarchical Flexgrid 案例:有时候我们需要将excel文件中数据采集出来存放入grid类型的控件进行显示,excel文件显示如下: wincc页面放置一个按钮,用于弹出消息框显示行数;放置Microsoft Hierarchical Flexgrid控件,设置名称为MSHFlex。wincc新建一个内部变量path,用于存放excel文件的路径。在wincc页面打开事件中写入一下VBS脚本。 Sub OnOpen() Dim path Set path=hmiruntime.Tags("path") path.writeHMIRuntime.ActiveProject.Path& "\myxls.xlsx" End Sub 在按钮的点击事件中写入以下脚本: Sub OnClick(ByVal Item) Dim objflex,xlApp,xlBook,path,RowCount,ColCount Dim i,j Set objflex=ScreenItems("MSHFlex") path=HMIRuntime.Tags("path").Read Set xlApp=CreateObject("excel.application") xlApp.Visible=False xlApp.Workbooks.Open path xlApp.Worksheets("Sheet1").Activate RowCount=xlApp.Worksheets("Sheet1").usedrange.rows.count ColCount=xlApp.Worksheets("Sheet1").usedrange.Columns.count objflex.Cols=ColCount+1 objflex.Rows=RowCount For i=1 To ColCount objflex.TextMatrix(0,i)=xlApp.Worksheets("Sheet1").cells(1,i).value Next For i=2 To RowCount For j=1 To ColCount objflex.TextMatrix(i-1,j)=xlApp.Worksheets("Sheet1").cells(i,j).value Next Next xlApp.Workbooks.Close xlApp.Quit

HierarchicalFlexGrid控件

HierarchicalFlexGrid控件 访问Hierarchical FlexGrid 控件 要在Visual Basic 中安装并访问Hierarchical FlexGrid 控件,请使用以下步骤。 要安装和访问Hierarchical FlexGrid 控件 1. 在“工程”菜单中,选择“部件”。出现“部件”对话框。 2. 在“控件”选项卡中,选择“Microsoft Hierarchical FlexGrid Control 6.0”,然后单击“确定”。MSHFlexGrid 控件被添加到Visual Basic 工具箱中。 3. 在Visual Basic 工具箱中,单击MSHFlexGrid 控件,然后将其拖到一个Visual Basic 窗体上。 -或者- 在Visual Basic 工具箱上,双击MSHFlexGrid 控件,将其添加到窗体上。 将数据绑定到Hierarchical FlexGrid 在开始使用它的功能之前,必须先将数据绑定到Hierarchical FlexGrid。要将数据绑定到控件,可以使用Visual Basic 新的Data Binding Manager,或者通过编程实现。 在将Hierarchical FlexGrid 绑定到数据源之后,Hierarchical FlexGrid 在设计时屏幕显示是一个空白列和一个空白行。字段和带区信息不是自动提取的(要获得此类信息,请参阅取得结构信息)。如果Hierarchical FlexGrid 在没有字段和带区信息的情况下运行,那么在显示数据的时候将使用缺省的属性设置。就是说,如果Hierarchical FlexGrid 被绑定到一个分层结构的Command,那么显示出的数据带区将是水平排列的,每个带区中包含一列,分别对应于Recordset 中的每一个字段。 绑定到数据源的Hierarchical FlexGrid 使用Visual Basic Data Binding Manager 将数据绑定到Hierarchical FlexGrid 本节说明如何使用Visual Basic Data Binding Manager 将数据绑定到Hierarchical FlexGrid。Data Binding Manager 提供了一种便于进行数据绑定的用户界面。 使用Visual Basic Data Binding Manager 设置DataSource 1. 为Hierarchical FlexGrid 创建数据源。数据源可以是DataEnvironment 对象或者ActiveXData Control,或者是一种新的Visual Basic 功能。在本例中,将数据源创建为DataEnvironment 对象。 2. 在Visual Basic 工具箱上,单击MSHFlexGrid 控件,然后将其拖到一个Visual Basic 窗体上。 -或者- 在Visual Basic 工具箱上,双击MSHFlexGrid 控件将其拖到一个VisualBasic 窗体上。 3. 在Visual Basic “属性”窗口中,将DataSource 属性设置为包含了希望被绑定到Hierarchical FlexGrid 的Command 对象的DataEnvironment对象。 警告如果DataSource 被重新设置,Hierarchical FlexGrid 单元格中的所有用户定义的、修改过的数据都将被丢失。4. 在Visual Basic “属性”窗口中,将DataMember 属性设置为DataEnvironment 中的一个Command 对象。如果希望在HierarchicalFlexGrid 中查看分层结构的数据,那么必须指定Command 分层结构中最顶部的父Command 对象作为

聚类(2)——层次聚类 Hierarchical Clustering .

聚类(2)——层次聚类Hierarchical Clustering 分类:Machine Learning 2012-06-23 11:09 5708人阅读评论(9) 收藏举报算法2010 聚类系列: ?聚类(序)----监督学习与无监督学习 ? ?聚类(1)----混合高斯模型 Gaussian Mixture Model ?聚类(2)----层次聚类 Hierarchical Clustering ?聚类(3)----谱聚类 Spectral Clustering -------------------------------- 不管是GMM,还是k-means,都面临一个问题,就是k的个数如何选取?比如在bag-of-words模型中,用k-means 训练码书,那么应该选取多少个码字呢?为了不在这个参数的选取上花费太多时间,可以考虑层次聚类。 假设有N个待聚类的样本,对于层次聚类来说,基本步骤就是: 1、(初始化)把每个样本归为一类,计算每两个类之间的距离,也就是样本与样本之间的相似度; 2、寻找各个类之间最近的两个类,把他们归为一类(这样类的总数就少了一个); 3、重新计算新生成的这个类与各个旧类之间的相似度; 4、重复2和3直到所有样本点都归为一类,结束。 整个聚类过程其实是建立了一棵树,在建立的过程中,可以通过在第二步上设置一个阈值,当最近的两个类的距离大于这个阈值,则认为迭代可以终止。另外关键的一步就是第三步,如何判断两个类之间的相似度有不少种方法。这里介绍一下三种: SingleLinkage:又叫做nearest-neighbor ,就是取两个类中距离最近的两个样本的距离作为这两个集合的距离,也就是说,最近两个样本之间的距离越小,这两个类之间的相似度就越大。容易造成一种叫做Chaining 的效果,两个cluster 明明从“大局”上离得比较远,但是由于其中个别的点距离比较近就被合并了,并且这样合并之后Chaining 效应会进一步扩大,最后会得到比较松散的cluster 。 CompleteLinkage:这个则完全是Single Linkage 的反面极端,取两个集合中距离最远的两个点的距离作为两个集合的距离。其效果也是刚好相反的,限制非常大,两个cluster 即使已经很接近了,但是只要有不配合的点存在,就顽固到底,老死不相合并,也是不太好的办法。这两种相似度的定义方法的共同问题就是指考虑了某个有特点的数据,而没有考虑类内数据的整体特点。 Average-linkage:这种方法就是把两个集合中的点两两的距离全部放在一起求一个平均值,相对也能得到合适一点的结果。 average-linkage的一个变种就是取两两距离的中值,与取均值相比更加能够解除个别偏离样本对结果的干扰。 这种聚类的方法叫做agglomerative hierarchical clustering(自下而上,@2013.11.20 之前把它写成自顶而下了,我又误人子弟了。感谢4楼的网友指正)的,描述起来比较简单,但是计算复杂度比较高,为了寻找距离最近/远和均值,都需要对所有的距离计算个遍,需要用到双重循环。另外从算法中可以看出,每次迭代都只能合并两个子类,这是非常慢的。尽管这么算起来时间复杂度比较高,但还是有不少地方用到了这种聚类方法,在《数学之美》一书的第14章介绍新闻分类的时候,就用到了自顶向下的聚类方法。 是这样的,谷歌02年推出了新闻自动分类的服务,它完全由计算机整理收集各个网站的新闻内容,并自动进行分类。新闻的分类中提取的特征是主要是词频因为对不同主题的新闻来说,各种词出现的频率是不一样的,比如科技报道类的新闻很可能出现的词就是安卓、平板、双核之类的,而军事类的新闻则更可能出现钓鱼岛、航

脑皮层的层次结构和Hierarchical Temporal Memory

?自然图像的统计特性和视网膜神经元的计算模型?脑皮层的层次结构和Hierarchical Temporal Memory ?Boltzmann机学习算法和脑的统计热力学模型

新皮层生理?大脑皮层中最晚进化出的 部分 ?特有的6层结构,在所有哺乳动物的大脑中发现 ?大脑中与智能有关的最主要部分(另外两个部分: 丘脑和海马) ?厚度:6张扑克牌 ?面积 –人:1张餐巾 –猴子:1个信封 –老鼠:1张邮票

新皮层分区和 层级结构?分区之间的层次关系 –例如,视觉腹侧 通路:Retina -> LGN -> V1 -> V2 -> V4 -> IT –层次关系通过分 区间的正向和反 向连接体现

关于新皮层的猜想 ?新皮层的结构在不同区域、不同物种之间,存在高度相似性 –层数相同、神经元种类和分布相同、连接方式相同?结构相似导致功能相似 –将初生雪雕的视觉信号连接到听觉发育区,听觉区发 育为视觉区 –先天性盲人阅读盲文时,激活的是视觉区 ?猜想(美国神经科学家Montcastle于1978年提出)–新皮层的功能区域的信息处理都遵循一个共同的算 法,视觉、听觉、运动输出等之间没有任何差异

耗散结构和生命体的区别 Benard花纹 生物体通过与环境交互,将自己限定在特定的 生态位中,维持自身的稳定结构。

脑的最基本功能是什么? ?避免surprise –Surprise对生命体通常意味着死亡 e.g., fish out of water –通过避免surprising state(减少熵),维持自身的存活(稳定状态) –Surprise是相对的 ?手段 –记忆、推断、预测 –指导行动

Lesson_2_Subsystems_Hierarchical_simulation

第二节:子系统——分层仿真 一个子系统就像一个元件——它有一个图标,参数,以及输入和输出端口。可以用一组元件或者其它子系统来创建一个子系统。在布局时通过分组选择元件可以很容易地创建一个子系统见图1。 当不同层次有大量的元件时,子系统可以让你通过已有元件库不用编程就可以创建你自己的元件、组织不同层次等级的布局。 图1 分层仿真 这一节描述了如何利用第一节[发射——外部调制激光]中讲到的“子系统外部调制激光”创建一个子系统。通过这一节你将对子系统和元件库更熟悉。 载入样本文件 载入样本文件,执行以下步骤: 步骤操作 1 从菜单中选择open。 2 从安装目录下选择Program Files > Optiwave Software >OptiSystem 6 > samples > Lesson 1.osd 创建一个子系统

创建一个子系统,执行以下操作: 步骤操作 1 在顶层图中,选择想让子系统包含的原件元件 选择的元件周围会出现一个框。 2 右键点击选择的元件。 将出现如图24的菜单。 3 从菜单选项中选择Create Subsystem. 子系统将出现在一个玻璃状盒子里。当查看子系统内部结构时 子系统标签将出现在顶层的底部见图2。 注意:可以看到的不包括没有连接的元件。子系统不会额外增加连接可以看到的原件的端口。 图2 创建一个子系统——在顶层选择元件 查看子系统内部结构 查看子系统内部结构,执行如下操作: 步骤操作 1 选择子系统玻璃盒点击右键。 将出现一个菜单见图3 2 在菜单选项中选择Look Inside. 子系统打开并且一个子系统标签出现在顶层在子系统层窗口的底部见 图3

MicrosoftHierarchicalFlexGrid控件使用方法

Microsoft Hierarchical FlexGrid 控件使用方法使用 Microsoft Hierarchical FlexGrid 控件 网址: Microsoft Hierarchical FlexGrid (MSHFlexGrid) 和 Microsoft FlexGrid (MSFlexGrid) 控件以网格的形式显示 Recordset 数据,数据可以来自单个表或者多个表。 Hierarchical FlexGrid 控件提供了在网格中显示数据的高级功能。它与Microsoft Data Bound 网格 (DataGrid) 控件类似,但也有显著区 别:Hierarchical FlexGrid 控件不允许用户对它绑定或包含的数据进行编辑。因此,这种控件在显示数据的同时能够确保原始数据的安全,使数据不被用户修改。不过,通过将它与文本框结合起来使用,Hierarchical FlexGrid 控件的单元格编辑能力也是可以实现的。尽管 Hierarchical FlexGrid 控件是以 Visual Basic 5.0 中使用的 FlexGrid 控件,Hierarchical FlexGrid 控件是比较灵活的。Hierarchical FlexGrid 控件还提供了更多的显示选项,利用这些选项我们可以定义出最适合于自己需要的自定义格式。这里的各个主题主要击中在如何使用Hierarchical FlexGrid 上。关于早期的 FlexGrid 控件的详细信息,请参阅有关的 Visual Basic 5.0 文档。 Visual Basic 的 FlexGrid 控件 图标缩写控件名称 MSHFlexGrid 控件 Microsoft Hierarchical FlexGrid 控件 MSFlexGrid 控件 Microsoft FlexGrid 控件

SPSS教程:Hierarchical Cluster分类分析

SPSS教程:Hierarchical Cluster分类分析 第二节 Hierarchical Cluster过程 10.2.1 主要功能 调用此过程可完成系统聚类分析。在系统聚类分析中,用户事先无法确定类别数,系统将所有例数均调入内存,且可执行不同的聚类算法。系统聚类分析有两种形式,一是对研究对象本身进行分类,称为Q型举类;另一是对研究对象的观察指标进行分类,称为R型聚类。 10.2.2 实例操作 [例10.2]29名儿童的血红蛋白(g/100ml)与微量元素(μg/100ml)测定结果如下表。由于微量元素的测定成本高、耗时长,故希望通过聚类分析(即R型指标聚类)筛选代表性指标,以便更经济快捷地评价儿童的营养状态。

10.2.2.1 数据准备 激活数据管理窗口,定义变量名:钙、镁、铁、锰、铜和血红蛋白的变量名分别为x1、x2、x3、x4、x5、x6,之后输入原始数据。 10.2.2.2 统计分析 激活Statistics 菜单选Classify 中的Hierarchical Cluster...项,弹出Hierarchical Cluster Analysis 对话框(图10.3)。从对话框左侧的变量列表中选x1、x2、x3、x4、x5、x6,点击 钮使之进入Variable(s)框;在Cluster 处选择聚类类型,其中Cases 表示观察对象聚类,Variables 表示变量聚类,本例选择Variables 。 图10.3 系统聚类分析对话框 点击Statistics...钮,弹出Hierarchical Cluster Analysis: Statistics 对话框,选择Distance matrix ,要求显示距离矩阵,点击Continue 钮返回Hierarchical Cluster Analysis 对话框(图10.4)。 图10.4 系统聚类方法选择对话框

设备间通信分层博弈Hierarchical Cooperation for Operator-Controlled device-to-device communications

a r X i v :1501.00620v 1 [c s .N I ] 4 J a n 2015 Hierarchical Cooperation for Operator-Controlled Device-to-Device Communications:A Layered Coalitional Game Approach Xiao Lu,Ping Wang,Dusit Niyato School of Computer Engineering,Nanyang Technological University,Singapore Email:{Luxiao,Wangping,Dniyato }@https://www.360docs.net/doc/ea5678829.html,.sg Abstract —Device-to-Device (D2D)communications,which al-low direct communication among mobile devices,have been proposed as an enabler of local services in 3GPP LTE-Advanced (LTE-A)cellular networks.This work investigates a hierarchical LTE-A network framework consisting of multiple D2D operators at the upper layer and a group of devices at the lower layer.We propose a cooperative model that allows the operators to improve their utility in terms of revenue by sharing their devices,and the devices to improve their payoff in terms of end-to-end throughput by collaboratively performing multi-path routing.To help understanding the interaction among operators and devices,we present a game-theoretic framework to model the cooperation behavior,and further,we propose a layered coalitional game (LCG)to address the decision making problems among them.Speci?cally,the cooperation of operators is modeled as an overlapping coalition formation game (CFG)in a partition form,in which operators should form a stable coalitional structure.Moreover,the cooperation of devices is modeled as a coalitional graphical game (CGG),in which devices establish links among each other to form a stable network structure for multi-path routing.We adopt the extended recursive core,and Nash network,as the stability concept for the proposed CFG and CGG,respectively.Numerical results demonstrate that the proposed LCG yields notable gains compared to both the non-cooperative case and a LCG variant and achieves good convergence speed. Keywords-D2D communications,LTE-Advanced network,layered coalitional game,coalitional structure formation,multi-path routing,extended recursive core,coalitional graph-ical game,Nash network. I.I NTRODUCTION Device-to-Device (D2D)communications [1],[2]have emerged as a promising paradigm for 3GPP LTE-Advanced (LTE-A)networks,which provide mobile wireless connectiv-ity,recon?gurable architectures,as well as various wireless applications (e.g.,network gaming,social content sharing and vehicular networking)for better user experience.With D2D communications,nearby devices in a cellular network can communicate with each other directly bypassing the base stations.Conventional D2D communications commonly refer to direct information exchanges among devices in Human-to-Human and Machine-to-Machine communications,without the involvement of wireless operators.However,conventional D2D technologies cannot provide ef?cient interference man-agement,security control and quality-of-service guarantee [2].Recently,there is a trend towards operator-controlled D2D communications to facilitate pro?t making for operators as well as better user experience for devices [2]. This paper considers a multi-hop LTE-A network consists of devices deployed by multiple operators.In this network,an ef?cient approach to improve the end-to-end throughput is to enable cooperative sharing of idle devices among the operators for multi-path routing [3].The cooperation can increase throughput for the devices because a cooperative relay may substantially lead to improved network capacity.Accordingly,a larger amount of user traf?c demand can be supported,which will lead to higher aggregated revenue for operators.In this cooperation,each operator needs to decide on which operators to cooperate with to maximize pro?t and,given the cooperation behavior of operators.Then,the devices from cooperative operators need to make decision on which devices to cooperate with to maximize throughput.We call the formation of this interrelated operator cooperation and device cooperation as a hierarchical cooperation problem,which is the main focus of this paper.The hierarchical cooperation gives rise to two major concerns.Firstly,what is the stable coalitional structure desirable for all operators so that none of operators is willing to leave the coalition?Secondly,what is the stable network structure for cooperative devices to perform multi-path routing?This paper addresses these two concerns by formulating a layered game framework to model the LTE-A network with operators and devices being the players in the upper and lower layer,respectively.Previous work has also considered game-theoretic framework with hierarchies/layers,e.g.,[4]in the cognitive radio networks,[5]in two-tier femtocell networks and [6]in WRNs.However,most of the works considered the competition relationship between differ-ent layers,which belongs to the Stackelberg game concept [7].In the proposed layered game framework,different layers interact to improve the bene?t of each other cooperatively.We adopt the concepts of an extended recursive core and Nash network as the solutions for the proposed games in the upper layer and lower layer,respective.To the best of our knowledge,this is the ?rst work to introduce the application of extended recursive core in wireless communications. II.N ETWORK M ODEL AND P ROBLEM F ORMULATION We consider an LTE-A network consisting of a number of devices belonging to multiple operators.We denote the set of operators as H ={1,2,...,H },and the set of devices of operator h ∈H as M (h )={1,2,...,M h }.The operators are willing to form overlapping coalitions to maximize their individual pro?ts.An overlapping coalitional structure for a number of operators can be de?ned in a cover function as the set πH ={S 1,S 2,...,S z }which is a collection of non-empty subsets of H such that z k =1S k =H ,?k,S k ?H and the sub-coalitions could overlap with each other.z is the total number of coalitions in collection πH .Let Γh denote the set of coalitions that operator h belongs to. For multiple access at every hop,we consider an OFDMA-

UHVDC分极分层接入方式及其运行特性

2018年2月电工技术学报Vol.33 No. 4 第33卷第4期TRANSACTIONS OF CHINA ELECTROTECHNICAL SOCIETY Feb. 2018 DOI: 10.19595/https://www.360docs.net/doc/ea5678829.html,ki.1000-6753.tces.161125 UHVDC分极分层接入方式及其运行特性 王玲1文俊1司瑞华2蔚泽1刘连光1 (1. 华北电力大学电气与电子工程学院北京 102206 2. 国网河南省电力公司经济技术研究院郑州 450000) 摘要在现有分层接入方式的基础上,提出分极分层接入方式。首先对特高压直流输电接入受端电网的两种分层接入方式,从经济性、多馈入短路比指标方面进行对比分析。继而基于 PSCAD/EMTDC仿真平台,对分层接入500kV/1 000kV受端电网的±800kV特高压直流输电系统 搭建了两种分层接入方式的仿真系统。最后对两种分层接入方式的特高压直流输电系统的稳态、 动态及暂态特性进行比较分析。研究结果表明:分极分层接入方式不仅能有效提高受端电网的多 馈入短路比,增强受端电网的电压支撑能力,而且具有良好的稳态及动态响应,更重要的是能够 实现两个逆变站换相失败的分层隔离,提高特高压直流输电运行的可靠性。文中针对分极分层接 入方式中单极停运或双极功率不平衡时接地极电流过大的不足,提出了抑制接地极电流的措施。 关键词:特高压直流分极分层接入方式多馈入短路比换相失败分层隔离故障恢复 中图分类号:TM72 The Connection Mode and Operation Characteristics of UHVDC with Hierarchical Connection by Pole Wang Ling1Wen Jun1 Si Ruihua2Yu Ze1 Liu Lianguang1 (1. School of Electrical and Electronic Engineering North China Electric Power University Beijing 102206 China 2. Henan Electrical Power Company Economic Research Institute Zhengzhou 450000 China) Abstract Based on the existing hierarchical connection mode, this paper puts forward a new hierarchical connection mode by pole. Firstly, according to the two hierarchical connection modes of ultra high voltage direct current (UHVDC) transmission system connecting to the receiving grid. the economic, multi-infeed short circuit ratio indexes are compared and analyzed. Then based on the PSCAD/EMTDC simulation platform, the simulation systems for 800kV UHVDC transmission system connecting to 500kV/1 000kV receiving grid are built. Finally, the steady-state, dynamic and transient characteristics of two connection modes are compared. The results show that the hierarchical connection mode by pole can not only effectively improve the multi-infeed short circuit ratio (MISCR) of receiving grid and enhance the voltage supporting ability, but also has good static and dynamic responses. More importantly, the hierarchical connection mode by pole can achieve the layered isolation of commutation failure and improve the reliability of UHVDC. The disadvantage of hierarchical connection mode by pole is that grounding current is too large when monopolar total outage or bipolar power is unbalanced. In this regard, this paper proposes the measures to suppress ground electrode current. Keywords:Ultra high voltage direct current (UHVDC), hierarchical connection by pole, multi-infeed short circuit ratio (MISCR), commutation failure, layered isolation, fault recovery 国家自然科学基金(51577060)和高等学校学科创新引智计划(“111”计划)资助项目。 收稿日期 2016-07-19 改稿日期 2016-10-11 万方数据

Hierarchical clustering

Statistics Toolbox Hierarchical Clustering Hierarchical clustering is a way to investigate grouping in your data, simultaneously over a variety of scales, by creating a cluster tree. The tree is not a single set of clusters, but rather a multi-level hierarchy, where clusters at one level are joined as clusters at the next higher level. This allows you to decide what level or scale of clustering is most appropriate in your application. The following sections explore the hierarchical clustering features in the Statistics Toolbox: Terminology and Basic Procedure Finding the Similarities Between Objects Defining the Links Between Objects Evaluating Cluster Formation Creating Clusters Terminology and Basic Procedure To perform hierarchical cluster analysis on a data set using the Statistics Toolbox functions, follow this procedure: Find the similarity or dissimilarity between every pair of objects in the data set. In this 1. step, you calculate the distance between objects using the pdist function. The pdist function supports many different ways to compute this measurement. See Finding the Similarities Between Objects for more information. 2. Group the objects into a binary, hierarchical cluster tree. In this step, you link together pairs of objects that are in close proximity using the linkage function. The linkage function uses the distance information generated in step 1 to determine the proximity of objects to each other. As objects are paired into binary clusters, the newly formed clusters are grouped into larger clusters until a hierarchical tree is formed. See Defining the Links Between Objects for more information. Determine where to divide the hierarchical tree into clusters. In this step, you divide the 3. objects in the hierarchical tree into clusters using the cluster function. The cluster function can create clusters by detecting natural groupings in the hierarchical tree or by cutting off the hierarchical tree at an arbitrary point. See Creating Clusters for more information. The following sections provide more information about each of these steps. Note The Statistics Toolbox includes a convenience function, clusterdata, which performs all these steps for you. You do not need to execute the pdist, linkage, or cluster functions separately. However, the clusterdata function does not give you access to the options each of the individual routines offers. For example, if you use the pdist function you can choose the distance calculation method, whereas if you use the clusterdata function you cannot. Finding the Similarities Between Objects

相关主题
相关文档
最新文档