• lgz
    了解作者
  • matlab
    开发工具
  • 301KB
    文件大小
  • zip
    文件格式
  • 0
    收藏次数
  • 1 积分
    下载积分
  • 10
    下载次数
  • 2018-12-07 10:38
    上传日期
将模板匹配与最近邻算法相结合的稳态负荷分解方法
Steady State Load Decomposition Method .zip
  • Steady State Load Decomposition Method Combining Template Matching with K-Nearest Neighbor Algorithms.pdf
    372KB
内容介绍
<html xmlns="http://www.w3.org/1999/xhtml"> <head> <meta charset="utf-8"> <meta name="generator" content="pdf2htmlEX"> <meta http-equiv="X-UA-Compatible" content="IE=edge,chrome=1"> <link rel="stylesheet" href="https://static.pudn.com/base/css/base.min.css"> <link rel="stylesheet" href="https://static.pudn.com/base/css/fancy.min.css"> <link rel="stylesheet" href="https://static.pudn.com/prod/directory_preview_static/626557414f8811599e787cdf/raw.css"> <script src="https://static.pudn.com/base/js/compatibility.min.js"></script> <script src="https://static.pudn.com/base/js/pdf2htmlEX.min.js"></script> <script> try{ pdf2htmlEX.defaultViewer = new pdf2htmlEX.Viewer({}); }catch(e){} </script> <title></title> </head> <body> <div id="sidebar" style="display: none"> <div id="outline"> </div> </div> <div id="pf1" class="pf w0 h0" data-page-no="1"><div class="pc pc1 w0 h0"><img class="bi x0 y0 w1 h1" alt="" src="https://static.pudn.com/prod/directory_preview_static/626557414f8811599e787cdf/bg1.jpg"><div class="t m0 x1 h2 y1 ff1 fs0 fc0 sc0 ls0 ws0">Steady St<span class="_ _0"></span>ate Load D<span class="_ _0"></span>ecom<span class="_ _0"></span>position Met<span class="_ _0"></span>h<span class="ls1 ws1">od Com<span class="_ _0"></span>bining Tem<span class="_ _0"></span>plate Matching with K-</span></div><div class="t m0 x2 h2 y2 ff1 fs0 fc0 sc0 ls1 ws1">Nearest Neighb<span class="_ _0"></span>or Algorithm<span class="_ _0"></span>s </div><div class="t m0 x3 h3 y3 ff2 fs1 fc0 sc0 ls2 ws2"> </div><div class="t m0 x4 h4 y4 ff2 fs2 fc0 sc0 ls3 ws3">Xiaoyu Z<span class="_ _0"></span>hang, Zhaowen Luan, Z<span class="_ _0"></span>hiliang Z<span class="_ _0"></span>hang </div><div class="t m0 x5 h3 y5 ff2 fs1 fc0 sc0 ls4 ws4">School of Electrical Engineering </div><div class="t m0 x6 h3 y6 ff2 fs1 fc0 sc0 ls4 ws5">Shandong University </div><div class="t m0 x7 h3 y7 ff2 fs1 fc0 sc0 ls5 ws6">Jina<span class="_ _0"></span>n, Ch<span class="_ _0"></span>ina </div><div class="t m0 x8 h3 y8 ff2 fs1 fc0 sc0 ls6 ws7">e-mail: 602336154@<span class="_ _0"></span>qq.com;<span class="_ _0"></span> luan-zw@sdu.edu.cn; 9353011<span class="_ _0"></span>16@qq.com </div><div class="t m0 x3 h3 y9 ff2 fs1 fc0 sc0 ls2 ws2"> </div><div class="t m0 x3 h3 ya ff2 fs1 fc0 sc0 ls2 ws2"> </div><div class="t m0 x9 h5 yb ff3 fs3 fc0 sc0 ls2 ws2">Abstract<span class="ff4">&#8212;<span class="ff1 ls7 ws8">Non-intrusive load monitoring a<span class="_ _1"></span>nd decomposition </span></span></div><div class="t m0 x9 h5 yc ff1 fs3 fc0 sc0 ls8 ws9">technology<span class="_ _1"></span> is one of t<span class="_ _1"></span>he most important c<span class="_ _1"></span>omponents<span class="_ _1"></span> of smart </div><div class="t m0 x9 h5 yd ff1 fs3 fc0 sc0 ls9 wsa">grid. It is the basis for in<span class="_ _0"></span>-depth analysis of user electricity </div><div class="t m0 x9 h5 ye ff1 fs3 fc0 sc0 lsa wsb">information, which is of great importance to improve user </div><div class="t m0 x9 h5 yf ff1 fs3 fc0 sc0 lsa wsc">quality of life an<span class="_ _0"></span>d improve grid serv<span class="_ _0"></span>ices. In this paper, a ne<span class="_ _0"></span>w </div><div class="t m0 x9 h5 y10 ff1 fs3 fc0 sc0 lsb wsd">load decomposition method combining te<span class="_ _1"></span>mplate matching w<span class="_ _1"></span>ith </div><div class="t m0 x9 h5 y11 ff1 fs3 fc0 sc0 lsc wse">k-nearest neighbor algorithm is proposed based on the study of </div><div class="t m0 x9 h5 y12 ff1 fs3 fc0 sc0 lsb wsf">steady-state load decomposition <span class="_ _1"></span>model. It solves the ex<span class="_ _1"></span>isting </div><div class="t m0 x9 h5 y13 ff1 fs3 fc0 sc0 lsd ws10">proble<span class="_ _1"></span>ms of <span class="_ _1"></span>unknow<span class="_ _1"></span>n l<span class="_ _1"></span>oad, de<span class="_ _1"></span>penden<span class="_ _1"></span>ce on trai<span class="_ _1"></span>ning<span class="_ _1"></span> data, <span class="_ _1"></span>a<span class="_ _1"></span>nd </div><div class="t m0 x9 h5 y14 ff1 fs3 fc0 sc0 lse ws11">difficult identification of low-pow<span class="_ _1"></span>er devices. It uses the </div><div class="t m0 x9 h5 y15 ff1 fs3 fc0 sc0 lsc ws12">collected data of the actual operation of the electrical </div><div class="t m0 x9 h5 y16 ff1 fs3 fc0 sc0 lsf ws13">appliances and through adding unknown load of different </div><div class="t m0 x9 h5 y17 ff1 fs3 fc0 sc0 lsb ws14">power, the simulation w<span class="_ _1"></span>as carried out by MATLAB<span class="_ _1"></span> to verify </div><div class="t m0 x9 h5 y18 ff1 fs3 fc0 sc0 ls10 ws15">that the m<span class="_ _0"></span>ethod in this paper is fea<span class="_ _0"></span>sible in practical </div><div class="t m0 x9 h5 y19 ff1 fs3 fc0 sc0 ls11 ws2">appli<span class="_ _1"></span>cati<span class="_ _1"></span>ons.<span class="ff3 ls2"> </span></div><div class="t m0 xa h6 y1a ff3 fs3 fc0 sc0 ls12 ws2">Keywords<span class="ff5 ls2">&#8212;</span><span class="ls13 ws16">k-nearest neighbor; ha<span class="ls10 ws17">r<span class="_ _0"></span>monic analysis; <span class="_ _1"></span>load </span></span></div><div class="t m0 x9 h6 y1b ff3 fs3 fc0 sc0 ls2 ws18">decomposition; no<span class="_ _1"></span>n-intrusive; steady state </div><div class="t m0 xb h3 y1c ff2 fs1 fc0 sc0 ls14 ws2">I.<span class="ff6 ls2"> <span class="_ _2"> </span><span class="ff2">I<span class="fs4 ls15">NTRODUCTION <span class="_ _1"></span></span> </span></span></div><div class="t m0 xa h3 y1d ff2 fs1 fc0 sc0 ls16 ws19">The const<span class="_ _0"></span>ructi<span class="_ _0"></span>on of<span class="_ _0"></span> a smart g<span class="_ _0"></span>rid w<span class="_ _0"></span>ill prom<span class="_ _0"></span>ote th<span class="_ _0"></span>e </div><div class="t m0 x9 h3 y1e ff2 fs1 fc0 sc0 ls17 ws1a">developm<span class="_ _0"></span>ent of<span class="_ _0"></span> smart com<span class="_ _0"></span>mun<span class="_ _0"></span><span class="ls18 ws1b">ities and sm<span class="_ _0"></span>art cit<span class="_ _0"></span>ies, enhanc<span class="_ _0"></span>e </span></div><div class="t m0 x9 h3 y1f ff2 fs1 fc0 sc0 ls19 ws1c">people's in<span class="_ _0"></span>tellig<span class="_ _0"></span>ent living s<span class="_ _0"></span>tandards<span class="_ _0"></span>, makin<span class="_ _0"></span>g life more </div><div class="t m0 x9 h3 y20 ff2 fs1 fc0 sc0 ls1a ws1d">convenient, l<span class="_ _0"></span>ower-carb<span class="_ _0"></span>on and <span class="_ _0"></span>ec<span class="ls1b ws1e">onom<span class="_ _0"></span>ical [1].<span class="_ _0"></span> As one<span class="_ _0"></span> of the<span class="_ _0"></span> </span></div><div class="t m0 x9 h3 y21 ff2 fs1 fc0 sc0 ls1c ws1f">key techn<span class="_ _0"></span>ologies of<span class="_ _0"></span> smart <span class="_ _0"></span>grid, <span class="_ _0"></span>load m<span class="_ _0"></span>onitoring is <span class="_ _0"></span>an </div><div class="t m0 x9 h3 y22 ff2 fs1 fc0 sc0 ls1d ws20">import<span class="_ _0"></span>ant step to rea<span class="_ _0"></span>lize smart<span class="_ _0"></span> electrici<span class="_ _0"></span>ty<span class="_ _0"></span>. </div><div class="t m0 xa h3 y23 ff2 fs1 fc0 sc0 ls2 ws21">Tradition<span class="_ _0"></span>al loa<span class="_ _0"></span>d monitorin<span class="_ _0"></span>g adopts an in<span class="_ _0"></span>trusiv<span class="_ _0"></span>e way<span class="_ _0"></span>, in </div><div class="t m0 x9 h3 y24 ff2 fs1 fc0 sc0 ls2 ws22">which sensors <span class="_ _0"></span>are in<span class="_ _0"></span>stalled <span class="_ _0"></span>on e<span class="_ _0"></span>ach user's <span class="_ _0"></span>elect<span class="_ _0"></span>rical<span class="_ _0"></span> </div><div class="t m0 x9 h3 y25 ff2 fs1 fc0 sc0 ls1e ws23">equipm<span class="_ _0"></span>ent to record thei<span class="_ _0"></span>r usage. Moni<span class="_ _0"></span>toring<span class="_ _0"></span> data is accur<span class="_ _0"></span>ate </div><div class="t m0 x9 h3 y26 ff2 fs1 fc0 sc0 ls1d ws24">and reli<span class="_ _0"></span>able by<span class="_ _0"></span> this w<span class="_ _0"></span>ay, bu<span class="_ _0"></span>t it<span class="_ _0"></span> also h<span class="_ _0"></span>as many<span class="_ _0"></span> probl<span class="_ _0"></span>ems such </div><div class="t m0 x9 h3 y27 ff2 fs1 fc0 sc0 ls1d ws25">as poor pr<span class="_ _0"></span>oject o<span class="_ _0"></span>perability<span class="_ _0"></span>, hig<span class="_ _0"></span>h implem<span class="_ _0"></span>entation c<span class="_ _0"></span>ost and </div><div class="t m0 x9 h3 y28 ff2 fs1 fc0 sc0 ls1f ws26">low u<span class="_ _0"></span>ser acce<span class="_ _0"></span>ptance. <span class="_ _0"></span>Non-intru<span class="_ _0"></span>sive l<span class="_ _0"></span>oad m<span class="_ _0"></span>onitorin<span class="_ _0"></span>g w<span class="_ _0"></span>as </div><div class="t m0 x9 h3 y29 ff2 fs1 fc0 sc0 ls1d ws27">propose<span class="_ _0"></span>d by H<span class="_ _0"></span>art [<span class="_ _0"></span>2] in<span class="_ _0"></span> the <span class="_ _0"></span>1980s. <span class="_ _0"></span>The ess<span class="_ _0"></span>ence <span class="_ _0"></span>is loa<span class="_ _0"></span>d </div><div class="t m0 x9 h3 y2a ff2 fs1 fc0 sc0 ls20 ws28">decom<span class="_ _0"></span>positi<span class="_ _0"></span>on, w<span class="_ _0"></span>h<span class="_ _0"></span>ich is t<span class="_ _0"></span>o de<span class="_ _0"></span>compos<span class="_ _0"></span>e th<span class="_ _0"></span>e to<span class="_ _0"></span>tal <span class="_ _0"></span>load<span class="_ _0"></span> </div><div class="t m0 x9 h3 y2b ff2 fs1 fc0 sc0 ls21 ws29">inform<span class="_ _0"></span>ation of use<span class="_ _0"></span>rs into the inf<span class="_ _0"></span>ormation of<span class="_ _0"></span> each elect<span class="_ _0"></span>rical<span class="_ _0"></span> </div><div class="t m0 x9 h3 y2c ff2 fs1 fc0 sc0 ls1e ws2a">equipm<span class="_ _0"></span>ent, so<span class="_ _0"></span> as t<span class="_ _0"></span>o obtain<span class="_ _0"></span> the<span class="_ _0"></span> electr<span class="_ _0"></span>icity<span class="_ _0"></span> inform<span class="_ _0"></span>ation<span class="_ _0"></span> such a<span class="_ _0"></span>s </div><div class="t m0 x9 h3 y2d ff2 fs1 fc0 sc0 ls22 ws2b">ener<span class="_ _1"></span>gy consumption of consumers a<span class="_ _1"></span>nd the power </div><div class="t m0 x9 h3 y2e ff2 fs1 fc0 sc0 ls23 ws7">consum<span class="_ _0"></span>ption<span class="_ _0"></span> rules <span class="_ _0"></span>of us<span class="_ _0"></span>ers<span class="_ _0"></span>. </div><div class="t m0 xa h3 y2f ff2 fs1 fc0 sc0 ls16 ws2c">In 2012, Ming<span class="_ _0"></span> Dong [3] an<span class="_ _0"></span>d <span class="ls24 ws2d">others <span class="_ _0"></span>prop<span class="_ _0"></span>osed a<span class="_ _0"></span> matchin<span class="_ _0"></span>g </span></div><div class="t m0 x9 h3 y30 ff2 fs1 fc0 sc0 ls1b ws2e">meth<span class="_ _0"></span>od base<span class="_ _0"></span>d on event<span class="_ _0"></span> wi<span class="_ _0"></span>ndow<span class="_ _0"></span>, <span class="ls1f ws2f">taking edge,<span class="_ _0"></span> tren<span class="_ _0"></span>d, sequen<span class="_ _0"></span>ce<span class="_ _0"></span>, </span></div><div class="t m0 x9 h3 y31 ff2 fs1 fc0 sc0 ls25 ws30">durati<span class="_ _0"></span>on and phas<span class="_ _0"></span>e fe<span class="_ _0"></span>atures o<span class="_ _0"></span>f each d<span class="_ _0"></span>evice <span class="_ _0"></span>as start-sto<span class="_ _0"></span>p and<span class="_ _0"></span> </div><div class="t m0 x9 h3 y32 ff2 fs1 fc0 sc0 ls2 ws31">change as th<span class="_ _0"></span>e main l<span class="_ _0"></span>oad im<span class="_ _0"></span>prin<span class="_ _0"></span>ts and usin<span class="_ _0"></span>g th<span class="_ _0"></span>e sample<span class="_ _0"></span> data<span class="_ _0"></span> </div><div class="t m0 x9 h3 y33 ff2 fs1 fc0 sc0 ls24 ws32">to cons<span class="_ _0"></span>ider<span class="_ _0"></span> weight of<span class="_ _0"></span> m<span class="_ _0"></span>ulti-featu<span class="_ _0"></span>re re<span class="_ _0"></span>cogn<span class="_ _0"></span>ition t<span class="_ _0"></span>raining,<span class="_ _0"></span> bu<span class="_ _0"></span>t </div><div class="t m0 x9 h3 y34 ff2 fs1 fc0 sc0 ls1f ws33">the m<span class="_ _0"></span>ethod of<span class="_ _0"></span> any test<span class="_ _0"></span> da<span class="_ _0"></span>ta rec<span class="_ _0"></span>ogniti<span class="_ _0"></span>on is<span class="_ _0"></span> lim<span class="_ _0"></span>ited and i<span class="_ _0"></span>t </div><div class="t m0 x9 h3 y35 ff2 fs1 fc0 sc0 ls19 ws34">needs <span class="_ _0"></span>to s<span class="_ _0"></span>et up <span class="_ _0"></span>a data<span class="_ _0"></span>base f<span class="_ _0"></span>or every<span class="_ _0"></span> use<span class="_ _0"></span>rs, w<span class="_ _0"></span>hich leads<span class="_ _0"></span> to <span class="_ _0"></span>the </div><div class="t m0 x9 h3 y36 ff2 fs1 fc0 sc0 ls25 ws35">implem<span class="_ _0"></span>entation of<span class="_ _0"></span> high dif<span class="_ _0"></span>ficu<span class="_ _0"></span>lty. Liter<span class="_ _0"></span>ature [4] est<span class="_ _0"></span>ablished </div><div class="t m0 xc h3 y37 ff2 fs1 fc0 sc0 ls1f ws33">the pa<span class="_ _0"></span>ramete<span class="_ _0"></span>r matrix<span class="_ _0"></span> of st<span class="_ _0"></span>eady<span class="_ _0"></span> state <span class="_ _0"></span>characte<span class="_ _0"></span>ristics <span class="_ _0"></span>of all </div><div class="t m0 xc h3 y38 ff2 fs1 fc0 sc0 ls26 ws36">kinds <span class="_ _0"></span>of el<span class="_ _0"></span>ectri<span class="_ _0"></span>c e<span class="_ _0"></span>quipm<span class="_ _0"></span>ent fo<span class="_ _0"></span>r l<span class="_ _0"></span>oad <span class="_ _0"></span>decom<span class="_ _0"></span>positi<span class="_ _0"></span>on,<span class="_ _0"></span> but<span class="_ _0"></span> it <span class="_ _0"></span>did<span class="_ _0"></span> </div><div class="t m0 xc h3 y39 ff2 fs1 fc0 sc0 ls1c ws37">not deal<span class="_ _0"></span> w<span class="_ _0"></span>ith the e<span class="_ _0"></span>quipm<span class="_ _0"></span>ent that is n<span class="_ _0"></span>ot in<span class="_ _0"></span>volved in<span class="_ _0"></span> th<span class="_ _0"></span>e </div><div class="t m0 xc h3 y3a ff2 fs1 fc0 sc0 ls16 ws38">param<span class="_ _0"></span>eter m<span class="_ _0"></span>atrix,<span class="_ _0"></span> w<span class="_ _0"></span>hich greatly<span class="_ _0"></span> aff<span class="_ _0"></span>ects th<span class="_ _0"></span>e accu<span class="_ _0"></span>racy<span class="_ _0"></span> of<span class="_ _0"></span> </div><div class="t m0 xc h3 y3b ff2 fs1 fc0 sc0 ls1e ws39">identifica<span class="_ _0"></span>tion in<span class="_ _0"></span> this case. In<span class="_ _0"></span> [5], an improv<span class="_ _0"></span>ed geneti<span class="_ _0"></span>c </div><div class="t m0 xc h3 y3c ff2 fs1 fc0 sc0 ls1f ws3a">algorithm<span class="_ _0"></span> w<span class="_ _0"></span>as used to en<span class="_ _0"></span>code th<span class="_ _0"></span>e op<span class="_ _0"></span>erating s<span class="_ _0"></span>tatus<span class="_ _0"></span> of a<span class="_ _0"></span> loa<span class="_ _0"></span>d </div><div class="t m0 xc h3 y3d ff2 fs1 fc0 sc0 ls27 ws3b">in bin<span class="_ _0"></span>ary<span class="_ _0"></span> mode<span class="_ _0"></span>, w<span class="_ _0"></span>hich can<span class="_ _0"></span> di<span class="_ _0"></span>rectly<span class="_ _0"></span> det<span class="_ _0"></span>ermin<span class="_ _0"></span>e the<span class="_ _0"></span> oper<span class="_ _0"></span>ating<span class="_ _0"></span> </div><div class="t m0 xc h3 y3e ff2 fs1 fc0 sc0 ls25 ws3c">status of<span class="_ _0"></span> the lo<span class="_ _0"></span>ad. However, th<span class="_ _0"></span>e accu<span class="_ _0"></span>racy<span class="_ _0"></span> of th<span class="_ _0"></span>e </div><div class="t m0 xc h3 y3f ff2 fs1 fc0 sc0 ls28 ws3d">decom<span class="_ _0"></span>positi<span class="_ _0"></span>on r<span class="_ _0"></span>esult<span class="_ _0"></span> is <span class="_ _0"></span>too <span class="_ _0"></span>depen<span class="_ _0"></span>dent <span class="_ _0"></span>on th<span class="_ _0"></span>e his<span class="_ _0"></span>tori<span class="_ _0"></span>cal<span class="_ _0"></span> </div><div class="t m0 xc h3 y40 ff2 fs1 fc0 sc0 ls29 ws3e">exper<span class="_ _1"></span>ience of evolutiona<span class="_ _1"></span>ry threshold.<span class="_ _1"></span> In [6], a method based </div><div class="t m0 xc h3 y41 ff2 fs1 fc0 sc0 ls2a ws2">on <span class="fc1 ls28 ws3f">local m<span class="_ _0"></span>ean <span class="_ _0"></span>decom<span class="_ _0"></span>positi<span class="_ _0"></span>on<span class="fc0 ls2b ws40"> w<span class="_ _0"></span>as us<span class="_ _0"></span>ed to sepa<span class="_ _0"></span>rate the<span class="_ _0"></span> mix<span class="_ _0"></span>ed </span></span></div><div class="t m0 xc h3 y42 ff2 fs1 fc0 sc0 ls2c ws41">data, an<span class="_ _0"></span>d the<span class="_ _0"></span> m<span class="_ _0"></span>ixed signal<span class="_ _0"></span> was se<span class="_ _0"></span>para<span class="_ _0"></span>ted in<span class="_ _0"></span>to a sin<span class="_ _0"></span>gle l<span class="_ _0"></span>oad of<span class="_ _0"></span> </div><div class="t m0 xc h3 y43 ff2 fs1 fc0 sc0 ls1f ws42">electrici<span class="_ _0"></span>ty consu<span class="_ _0"></span>mption<span class="_ _0"></span> data, <span class="_ _0"></span>and l<span class="_ _0"></span>oad matchin<span class="_ _0"></span>g w<span class="_ _0"></span>as </div><div class="t m0 xc h3 y44 ff2 fs1 fc0 sc0 ls2d ws43">performed through model matching. H<span class="_ _0"></span>owever, the method </div><div class="t m0 xc h3 y45 ff2 fs1 fc0 sc0 ls2c ws44">requi<span class="_ _0"></span>res hig<span class="_ _0"></span>h da<span class="_ _0"></span>ta colle<span class="_ _0"></span>ction<span class="_ _0"></span>, an<span class="_ _0"></span>d the n<span class="_ _0"></span>oise ex<span class="_ _0"></span>ceeding th<span class="_ _0"></span>e </div><div class="t m0 xc h3 y46 ff2 fs1 fc0 sc0 ls2e ws45">General A<span class="_ _0"></span>ssembly<span class="_ _0"></span> to inf<span class="_ _0"></span>luence th<span class="_ _0"></span>e extr<span class="_ _0"></span>eme points<span class="_ _0"></span>, an<span class="_ _0"></span>d the<span class="_ _0"></span> </div><div class="t m0 xc h3 y47 ff2 fs1 fc0 sc0 ls17 ws46">length of th<span class="_ _0"></span>e <span class="_ _0"></span>data use<span class="_ _0"></span>d in<span class="ls18 ws47"> th<span class="_ _0"></span>e Model establishm<span class="_ _0"></span>ent and </span></div><div class="t m0 xc h3 y48 ff2 fs1 fc0 sc0 ls2c ws48">Model m<span class="_ _0"></span>atching has<span class="_ _0"></span> an im<span class="_ _0"></span>pact<span class="_ _0"></span> on th<span class="_ _0"></span>e rec<span class="_ _0"></span>ognition<span class="_ _0"></span> accura<span class="_ _0"></span>cy. </div><div class="t m0 xd h3 y49 ff2 fs1 fc0 sc0 ls16 ws49">In summ<span class="_ _0"></span>ary<span class="_ _0"></span>, the non-<span class="_ _0"></span>inva<span class="_ _0"></span>sive loa<span class="_ _0"></span>d dec<span class="_ _0"></span>omposi<span class="_ _0"></span>tion<span class="_ _0"></span> </div><div class="t m0 xc h3 y4a ff2 fs1 fc0 sc0 ls24 ws4a">method<span class="_ _0"></span> based<span class="_ _0"></span> on<span class="_ _0"></span> steady-<span class="_ _0"></span>state inf<span class="_ _0"></span>orm<span class="_ _0"></span>ation h<span class="_ _0"></span>as <span class="_ _0"></span>been<span class="_ _0"></span> great<span class="_ _0"></span>ly </div><div class="t m0 xc h3 y4b ff2 fs1 fc0 sc0 ls2 ws4b">developed. <span class="_ _0"></span>However,<span class="_ _0"></span> ther<span class="_ _0"></span>e are sti<span class="_ _0"></span>ll no<span class="_ _0"></span> mature s<span class="_ _0"></span>olutions <span class="_ _0"></span>to </div><div class="t m0 xc h3 y4c ff2 fs1 fc0 sc0 ls2f ws4c">the pro<span class="_ _0"></span>blems, fo<span class="_ _0"></span>r exam<span class="_ _0"></span>ple, un<span class="_ _0"></span>know<span class="_ _0"></span>n devices in <span class="_ _0"></span>the lo<span class="_ _0"></span>ad, the<span class="_ _0"></span> </div><div class="t m0 xc h3 y4d ff2 fs1 fc0 sc0 ls27 ws4d">character<span class="_ _0"></span>istics of<span class="_ _0"></span> load dec<span class="_ _0"></span>ompos<span class="_ _0"></span>ition a<span class="_ _0"></span>re not o<span class="_ _0"></span>bvious<span class="_ _0"></span>, and </div><div class="t m0 xc h3 y4e ff2 fs1 fc0 sc0 ls25 ws4e">the data ar<span class="_ _0"></span>e too dependent<span class="_ _0"></span> on training dat<span class="_ _0"></span>a. Theref<span class="_ _0"></span>ore, this<span class="_ _0"></span> </div><div class="t m0 xc h3 y4f ff2 fs1 fc0 sc0 ls20 ws38">paper p<span class="_ _0"></span>resents a re<span class="_ _0"></span>cognit<span class="_ _0"></span>ion m<span class="_ _0"></span>ethod based on<span class="_ _0"></span> templat<span class="_ _0"></span>e </div><div class="t m0 xc h3 y50 ff2 fs1 fc0 sc0 ls30 ws4f">matching<span class="_ _0"></span> and <span class="fc1 ls1d ws50">k-Nea<span class="_ _0"></span>rest Ne<span class="_ _0"></span>ighbor<span class="_ _0"></span> (KNN)<span class="fc0 ls18 ws51"> alg<span class="_ _0"></span>orithm<span class="_ _0"></span>, w<span class="_ _0"></span>hich </span></span></div><div class="t m0 xc h3 y51 ff2 fs1 fc0 sc0 ls2c ws52">can not <span class="_ _0"></span>only<span class="_ _0"></span> solve th<span class="_ _0"></span>e <span class="_ _0"></span>applicabil<span class="_ _0"></span>ity<span class="_ _0"></span> of tra<span class="_ _0"></span>diti<span class="_ _0"></span>onal ste<span class="_ _0"></span>ady-state<span class="_ _0"></span> </div><div class="t m0 xc h3 y52 ff2 fs1 fc0 sc0 ls31 ws53">current d<span class="_ _0"></span>ecom<span class="_ _0"></span>positi<span class="_ _0"></span>on algo<span class="_ _0"></span>rithm<span class="_ _0"></span> f<span class="_ _0"></span>or the identif<span class="_ _0"></span>icati<span class="_ _0"></span>on of<span class="_ _0"></span> </div><div class="t m0 xc h3 y53 ff2 fs1 fc0 sc0 ls32 ws54">unknown device load, <span class="ls33 ws55">but als<span class="_ _0"></span>o improve th<span class="_ _0"></span>e dependence of<span class="_ _0"></span> </span></div><div class="t m0 xc h3 y54 ff2 fs1 fc0 sc0 ls1d ws56">template<span class="_ _0"></span> matching on<span class="_ _0"></span> the sam<span class="_ _0"></span>ple data excessiv<span class="_ _0"></span>ely. In<span class="_ _0"></span> </div><div class="t m0 xc h3 y55 ff2 fs1 fc0 sc0 ls24 ws57">practic<span class="_ _0"></span>al a<span class="_ _0"></span>pplicat<span class="_ _0"></span>ion,<span class="_ _0"></span> the<span class="_ _0"></span> reliabi<span class="_ _0"></span>lity <span class="_ _0"></span>of th<span class="_ _0"></span>e de<span class="_ _0"></span>composit<span class="_ _0"></span>ion<span class="_ _0"></span> </div><div class="t m0 xc h3 y56 ff2 fs1 fc0 sc0 ls34 ws58">result is<span class="_ _0"></span> guarante<span class="_ _0"></span>ed. </div><div class="t m0 xe h3 y57 ff2 fs1 fc0 sc0 ls14 ws2">II.<span class="ff6 ls2"> <span class="_ _3"> </span><span class="ff2">B</span></span></div><div class="t m0 xf h3 y58 ff2 fs4 fc0 sc0 ls35 ws2">ASIC<span class="_ _1"></span> <span class="fs1 ls2">M<span class="_ _1"></span></span><span class="ls36">ODEL<span class="fs1 ls2"> </span></span></div><div class="t m0 xc h7 y59 ff7 fs1 fc0 sc0 ls37 ws2">A.<span class="ff8 ls2"> <span class="_"> </span></span><span class="ls23 ws7">Load Decomp<span class="_ _1"></span>osition Model </span></div><div class="t m0 xd h3 y5a ff2 fs1 fc0 sc0 ls1e ws59">Load stea<span class="_ _0"></span>dy-s<span class="_ _0"></span>tate current<span class="_ _0"></span> decom<span class="_ _0"></span>position<span class="_ _0"></span> is<span class="_ _0"></span> divide<span class="_ _0"></span>d int<span class="_ _0"></span>o </div><div class="t m0 xc h3 y5b ff2 fs1 fc0 sc0 ls38 ws5a">horiz<span class="_ _0"></span>ontal de<span class="_ _0"></span>com<span class="_ _0"></span>position<span class="_ _0"></span> and verti<span class="_ _0"></span>cal<span class="_ _0"></span> decom<span class="_ _0"></span>position<span class="_ _0"></span>. </div><div class="t m0 xc h3 y5c ff2 fs1 fc0 sc0 ls17 ws5b">Horizont<span class="_ _0"></span>al dec<span class="_ _0"></span>ompositi<span class="_ _0"></span>on re<span class="_ _0"></span>fers t<span class="_ _0"></span>o the <span class="_ _0"></span>total cu<span class="_ _0"></span>rrent i<span class="_ _0"></span>s </div><div class="t m0 xc h3 y5d ff2 fs1 fc0 sc0 ls39 ws5c">decomposed into a fundam<span class="_ _0"></span>ental <span class="ws5d">wave and the fo<span class="_ _0"></span>rm of the </span></div><div class="t m0 xc h3 y5e ff2 fs1 fc0 sc0 ls38 ws5e">addition<span class="_ _0"></span> of<span class="_ _0"></span> each h<span class="_ _0"></span>arm<span class="_ _0"></span>onic, w<span class="_ _0"></span>h<span class="_ _0"></span>ile the v<span class="_ _0"></span>erti<span class="_ _0"></span>cal <span class="_ _0"></span>decom<span class="_ _0"></span>positi<span class="_ _0"></span>on </div><div class="t m0 xc h3 y5f ff2 fs1 fc0 sc0 ls1e ws5f">refers t<span class="_ _0"></span>o the tot<span class="_ _0"></span>al current <span class="_ _0"></span>is div<span class="_ _0"></span>ide<span class="_ _0"></span>d into<span class="_ _0"></span> a sing<span class="_ _0"></span>le devi<span class="_ _0"></span>ce </div><div class="t m0 xc h3 y60 ff2 fs1 fc0 sc0 ls31 ws60">current su<span class="_ _0"></span>perp<span class="_ _0"></span>osition<span class="_ _0"></span> form<span class="_ _0"></span>. Am<span class="_ _0"></span>ong th<span class="_ _0"></span>em, ho<span class="_ _0"></span>rizont<span class="_ _0"></span>al </div><div class="t m0 xc h3 y61 ff2 fs1 fc0 sc0 ls3a ws61">decom<span class="_ _0"></span>positi<span class="_ _0"></span>on<span class="_ _0"></span> can be o<span class="_ _0"></span>btain<span class="_ _0"></span>ed direc<span class="_ _0"></span>tly fr<span class="_ _0"></span>om harm<span class="_ _0"></span>onic<span class="_ _0"></span> </div><div class="t m0 x10 h8 y62 ff9 fs3 fc0 sc0 ls2 ws2">250</div><div class="t m0 x11 h9 y63 ff9 fs5 fc0 sc0 ls2 ws62">2018 International Conference on Smart Grid and Clean Energy Technologies</div><div class="t m0 x9 ha y64 ff2 fs3 fc0 sc0 ls3b ws63">978-1-5386-6410-0/18/$31.00 &#169;2018 IEEE</div></div><div class="pi" data-data='{"ctm":[1.611639,0.000000,0.000000,1.611639,0.000000,0.000000]}'></div></div> </body> </html>
评论
    相关推荐
    • report.rar
      写作实验报告,大家可以用作模板,特此向大家共享啦
    • matlabcnhelp.rar
      matlab中文帮助很难找的,快速下载
    • MobilePolice.rar
      移动警察,车牌识别,车牌定位系统源代码,已经运用在移动车载稽查系统中。
    • SVM(matlab).rar
      支持向量机(SVM)实现的分类算法源码[matlab]
    • svm.zip
      用MATLAB编写的svm源程序,可以实现支持向量机,用于特征分类或提取
    • Classification-MatLab-Toolbox.rar
      模式识别matlab工具箱,包括SVM,ICA,PCA,NN等等模式识别算法,很有参考价值
    • VC++人脸定位实例.rar
      一个经典的人脸识别算法实例,提供人脸五官定位具体算法及两种实现流程.
    • QPSK_Simulink.rar
      QPSK的Matlab/Simulink的调制解调仿真系统,给出接收信号眼图及系统仿真误码率,包含载波恢复,匹配滤波,定时恢复等重要模块,帮助理解QPSK的系统
    • LPRBPDemo2009KV.rar
      车牌识别,神经网络算法,识别率高达95%,识别时间低于80ms。
    • MODULATION.RAR
      这个源程序代码包提供了通信系统中BPSK,QPSK,OQPSK,MSK,MSK2,GMSK,QAM,QAM16等调制解调方式 用matlab的实现,以及它们在AWGN和Rayleigh信道下的通信系统实现及误码率性能