OASIS Repository@POSTECHLIBRARY: 하이 패스 필터와 딥러닝을 이용한 협동로봇의 충돌감지

<!DOCTYPE html> <html lang="ko"> <head> <title>OASIS Repository@POSTECHLIBRARY: 하이 패스 필터와 딥러닝을 이용한 협동로봇의 충돌감지</title> <meta http-equiv="Content-Type" content="text/html; charset=UTF-8" /> <meta name="Generator" content="DSpace 5.5" /> <meta name="viewport" content="width=device-width, initial-scale=1.0"> <meta http-equiv="X-UA-Compatible" content="IE=edge"> <meta name="description" content="Postech OASIS Repository" /> <meta name="keywords" content="오아시스,레포지토리,리포지터리,저장소,포스텍,박태준학술정보관,포항공과대학교,포항공대,POSTECH,Pohang University of Science and Technology,도서관,IR,연구업적,학술실적,연구성과,dspace" /> <link rel="shortcut icon" href="/favicon.ico" type="image/x-icon"/> <link rel="search" type="application/opensearchdescription+xml" href="/open-search/description.xml" title="DSpace" /> <link rel="schema.DCTERMS" href="http://purl.org/dc/terms/" /> <link rel="schema.DC" href="http://purl.org/dc/elements/1.1/" /> <link rel="schema.OAK" href="http://www.oak.go.kr/terms/" /> <meta name="OAK.author" content="KIM, TO IL" scheme="OAK.AUTHOR" /> <meta name="OAK.author" content="CHUNG, WAN KYUN" scheme="OAK.AUTHOR" /> <meta name="DCTERMS.dateAccepted" content="2020-09-18T09:51:39Z" scheme="DCTERMS.W3CDTF" /> <meta name="DCTERMS.available" content="2020-09-18T09:51:39Z" scheme="DCTERMS.W3CDTF" /> <meta name="DCTERMS.created" content="2020-08-31" scheme="DCTERMS.W3CDTF" /> <meta name="DCTERMS.issued" content="2020-08-17" scheme="DCTERMS.W3CDTF" /> <meta name="DC.identifier" content="https://oasis.postech.ac.kr/handle/2014.oak/104146" scheme="DCTERMS.URI" /> <meta name="DCTERMS.abstract" content="Collision detection is an important problem for safe use of collaborative robots around humans. Although the generalized momentum observer is a widely used proprioceptive collision detection method, its performance is degraded by inaccuracy in dynamics and in friction modelling. A deep learning approach was recently proposed to overcome such limitations of the model-based generalized momentum observer, but its application is limited to fixed repetitive motions. In this paper, a deep learning approach is proposed which can also be applied to non-repetitive motions. Robot signals are passed through a first order high pass filter, then the filtered signals are used to train a convolutional neural network(CNN) to detect collisions. The performance of the trained network is evaluated in terms of false negative, false positive rate, and average detection time." /> <meta name="DC.language" content="Korean" /> <meta name="DC.publisher" content="한국로봇학회" /> <meta name="DC.relation" content="제 15회 한국로봇종합학술대회" /> <meta name="DC.relation" content="제 15회 한국로봇종합학술대회" /> <meta name="DC.title" content="하이 패스 필터와 딥러닝을 이용한 협동로봇의 충돌감지" /> <meta name="DCTERMS.alternative" content="Collision Detection of Collaborative Robot Using High Pass Filter and Deep Learning" /> <meta name="DC.type" content="Conference" /> <meta name="DC.identifier" content="57377" /> <meta name="DC.type" content="CONF" /> <meta name="DC.identifier" content="제 15회 한국로봇종합학술대회, pp.211 - 212" /> <meta name="OAK.relation.page" content="212" scheme="OAK.PAGE" /> <meta name="OAK.relation.page" content="211" scheme="OAK.PAGE" /> <meta name="OAK.relation.journal" content="제 15회 한국로봇종합학술대회" scheme="OAK.JOURNAL" /> <meta name="DC.contributor" content="KIM, TO IL" /> <meta name="DC.contributor" content="CHUNG, WAN KYUN" /> <meta name="DC.description" content="2" /> <meta name="DC.description" content="2" /> <meta name="citation_keywords" content="Conference" /> <meta name="citation_title" content="하이 패스 필터와 딥러닝을 이용한 협동로봇의 충돌감지" /> <meta name="citation_publisher" content="한국로봇학회" /> <meta name="citation_author" content="KIM, TO IL" /> <meta name="citation_author" content="CHUNG, WAN KYUN" /> <meta name="citation_date" content="2020-08-17" /> <meta name="citation_abstract_html_url" content="https://oasis.postech.ac.kr/handle/2014.oak/104146" /> <link rel="stylesheet" href="/css/bootstrap.min.css" defer /> <link rel="stylesheet" href="/css/layout.css" async /> <link rel="stylesheet" href="/css/mquery.css" defer /> <link rel="stylesheet" href="/css/slidebars.css" defer />  <link rel="stylesheet" href="/css/owl.carousel.css" defer />  <link rel="stylesheet" href="/css/owl.theme.css" defer /> <link rel="stylesheet" href="/css/bootstrap-partof.css" defer /> <script src="/js/jquery-1.9.1.min.js"></script> <script src="/js/jquery-ui.js"></script> <script src="/js/bootstrap.min.js"></script> <script src="/js/owl.carousel.min.js"></script> <script src="/js/common.js" defer></script> <script src="/utils.js" defer></script> <script src="/static/js/holder.js" defer></script> <script src="/static/js/choice-support.js" defer></script> <script src="/js/ms-clarity.js"></script> <script async src="https://www.googletagmanager.com/gtag/js?id=G-B9EHYYGM78"></script> <script> window.dataLayer = window.dataLayer || []; function gtag(){dataLayer.push(arguments);} gtag('js', new Date()); gtag('config', 'G-B9EHYYGM78'); </script>   </head> <body> <div id="sb-site"> <script> function doSearch () { (function($) { if ($("select[name='filtername']").val() != '') { if ($("input:text[name='query']").val() == '') { $("input:text[name='contains']").val(""); $("select[name='filtername']").attr("disabled", true); $("input:hidden[name='filtertype']").attr("disabled", true); $("input:hidden[name='filterquery']").attr("disabled", true); } else { $("input:hidden[name='filterquery']").val($("input:text[name='query']").val()); $("input:text[name='query']").val(""); } } else { $("input:text[name='contains']").val(""); $("select[name='filtername']").attr("disabled", true); $("input:hidden[name='filtertype']").attr("disabled", true); $("input:hidden[name='filterquery']").attr("disabled", true); } })(jQuery.noConflict()); } </script> <div class="col_width sub_header"> <h1><a href="/">Open Access System for Information Sharing</a></h1> <form action="/simple-search" method="get" onsubmit="doSearch();"> <input type="hidden" name="filtertype" value="contains" /> <div class="sub_search_box"> <span class="ss_select"> <select name="filtername" id="header_filter"> <option value="">All</option> <option value="title">Title</option> <option value="author">Author</option> <option value="subject">Subject</option> </select> </span> <div class="ss_int_box"> <input type="text" title="검색창" class="ms_int" name="query" placeholder="Enter Search keyword"/> <input type="hidden" name="filterquery"/> <input type="submit" title="검색" class="ms_bt" value="search" /> </div> </div> </form> <div class="gnav"> <a href="/password-login" class="first_a">Login</a> <a href="http://library.postech.ac.kr/" target="_blank">Library</a> <script type="text/javascript">  </script><noscript><a href="/help/index.html#" target="dspacepopup">Help</a></noscript></div> <div class="tablet_nav sb-toggle-right"> <a href="#" class="tablet_nav_bt" id="mnav_bt"> <span class="line"></span> <span class="line"></span> <span class="line"></span> </a> </div> <div class="mobile_search"> <form action="/simple-search" method="get" id="search_form"> <a href="javascript:doSearch($('#search_form'));" class="mobile_sbt">검색</a> <div class="mobile_s_inner"> <span class="d_arrow"></span> <p class="mmobile_s_int"><input type="text" title="검색" placeholder="Search"></p> </div> </form> </div> </div>  <div class="sub_nav_wrap"> <div class="col_width"> <div class="sub_nav_box"> <ul> <li class="home_nav"><a href="/" >HOME</a></li> <li><a href="/community-list" >Communities & Collections</a></li> <li><a href="/browse-researcher" >Researchers</a></li> <li><a href="/browse?type=title" class='on'>Title</a></li> </ul> </div> <div class="page_nav"> <a href="/" class="page_home"><strong> </strong></a> <a href="/handle/2014.oak/408" style="max-width: 170px;"><strong>Department of Mechanical Engineering (기계공학과)</strong></a> <a href="/handle/2014.oak/410" ><strong>2. Conference Papers</strong></a> </div></div> </div> <div class="col_width sub_container add_widget"> <div class="sub_contents"> <script src="https://apis.google.com/js/platform.js" async defer></script> <script type="text/javascript"> // google plus api window.___gcfg = {lang: 'ko'}; (function() { var po = document.createElement('script'); po.type = 'text/javascript'; po.async = true; po.src = 'https://apis.google.com/js/platform.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(po, s); })(); </script> <div class="sub_title"> <h3> </h3> <div class="type_icon"> <span class="article_type">Conference</span> </div> </div> <div class="view_top_box"> <div class="view_bt_box"> <span class="cited_span science_span">Cited <em>0</em> time in <img src="/image/common/webofscience.png" alt="webofscience"></span> <span class="cited_span">Cited <em>0</em> time in <img src="/image/common/scopus.png" alt="scopus"></span> </div> <div class="view_bt_area"> <span>Metadata Downloads</span> <form action="/export" method="post"> <input type="hidden" name="item_id" value="103808"/> <div class="view_downbt"> <ul> <li> <input type="submit" name="submit_export_dc" value="DC(XML)"/> </li> <li> <input type="submit" name="submit_export_excel" value="EXCEL"/> </li> </ul> </div> </form> </div> </div> <div class="view_contents"> <p class="view_title">하이 패스 필터와 딥러닝을 이용한 협동로봇의 충돌감지 </p> <div class="view_inner_con"> <dl><dt>Title</dt><dd>하이 패스 필터와 딥러닝을 이용한 협동로봇의 충돌감지</dd></dl> <dl><dt>Authors</dt><dd><a class="author"href="/browse?type=author&value=KIM%2C+TO+IL">KIM, TO IL</a>; <a class="author_a" href="/researcher-profile?ep=637">CHUNG, WAN KYUN</a></dd></dl> <dl><dt>Date Issued</dt><dd>2020-08-17 </dd></dl> <dl><dt>Publisher</dt><dd>한국로봇학회</dd></dl> <dl><dt>Abstract</dt><dd>Collision detection is an important problem for safe use of collaborative robots around humans. Although the generalized momentum observer is a widely used proprioceptive collision detection method, its performance is degraded by inaccuracy in dynamics and in friction modelling. A deep learning approach was recently proposed to overcome such limitations of the model-based generalized momentum observer, but its application is limited to fixed repetitive motions. In this paper, a deep learning approach is proposed which can also be applied to non-repetitive motions. Robot signals are passed through a first order high pass filter, then the filtered signals are used to train a convolutional neural network(CNN) to detect collisions. The performance of the trained network is evaluated in terms of false negative, false positive rate, and average detection time.</dd></dl> <dl><dt>URI</dt><dd><a href="https://oasis.postech.ac.kr/handle/2014.oak/104146" class="link_type">https://oasis.postech.ac.kr/handle/2014.oak/104146</a></dd></dl> <dl><dt>Article Type</dt><dd>Conference</dd></dl> <dl><dt>Citation</dt><dd>제 15회 한국로봇종합학술대회, page. 211 - 212, 2020-08-17</dd></dl> <dl class="file_item_dl"><dt>Files in This Item:</dt> <dd class="file_download">There are no files associated with this item.</dd> </dl> </div> <div class="record_bt_box"> <a href="/handle/2014.oak/104146?mode=full">Show full item record</a> </div> <div class="al_right"> </div> <div class="sns_wrap"> <div class="sns_box"> <p class="qr_box"><img src="https://api.qrserver.com/v1/create-qr-code/?size=66x66&data=https://oasis.postech.ac.kr/handle/2014.oak/104146" alt="qr_code"></p> <div class="sns_inner"> <ul> <li> <p><a href="http://www.mendeley.com/import/?url=https://oasis.postech.ac.kr/handle/2014.oak/104146" target="_blank"><img src="/image/common/mendeley_icon.gif" alt="mendeley" /></a></p> <p class="twitter_box"> <a href="https://twitter.com/share" class="tweet_bt twitter-share-button" data-lang="en" data-size="large" data-dnt="true">트윗하기</a> <script> !function(d, s, id) { var js, fjs = d.getElementsByTagName(s)[0]; if (!d.getElementById(id)) { js = d.createElement(s); js.id = id; js.src = "//platform.twitter.com/widgets.js"; fjs.parentNode.insertBefore(js, fjs); } }(document, "script", "twitter-wjs"); </script> </p> </li> <li class="facebook_li"> <span id="fb-root"></span> <script> (function(d, s, id) { var js, fjs = d.getElementsByTagName(s)[0]; if (d.getElementById(id)) return; js = d.createElement(s); js.id = id; js.src = "//connect.facebook.net/ko_KR/all.js#xfbml=1"; fjs.parentNode.insertBefore(js, fjs); }(document, 'script', 'facebook-jssdk')); </script> <span class="fb-like" data-send="true" data-layout="standard" data-width="450" data-show-faces="false" data-font="verdana"></span> </li> </ul> </div> </div> <p class="sns_text"> <span> Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.</span> </p> </div> </div> </div>  <div class="sub_right_box"> <div class="w_cc"> <h4 class="widget_title">Communities & Collection</h4> <ul> <li><a href="/handle/2014.oak/408" class="cc_item">Department of Mechanical Engineering (기계공학과)</a> <ul> <li><a href="/handle/2014.oak/409">1. Journal Papers<span class="round_num"><em>3,389</em></span></a></li> <li><a href="/handle/2014.oak/410">2. Conference Papers<span class="round_num"><em>6,447</em></span></a></li> <li><a href="/handle/2014.oak/412">3. Theses_Ph.D.<span class="round_num"><em>424</em></span></a></li> <li><a href="/handle/2014.oak/411">4. Theses_Master<span class="round_num"><em>823</em></span></a></li> <li><a href="/handle/2014.oak/9252">ETC<span class="round_num"><em>0</em></span></a></li> </ul> </li> </ul> </div> <div class="w_researcher"> <h4 class="widget_title">Related Researcher</h4> <div class="researcher_area researcher_area_detail"> <p class="researcher_img_box"><span><img src="" alt="Researcher" onerror="javascript:this.src='/image/common/no_img.gif'"></span></p> <div class="reseacher_info"> <dl> <dt><a href="/researcher-profile?ep=637">정완균<span>CHUNG, WAN KYUN</span></a></dt> <dd class="interests_dd">Dept of Mechanical Enginrg</dd> </dl> <a href="/researcher-profile?ep=637" class="read_more"><em>Read more</em></a> </div> </div> </div> <script type='text/javascript' src='https://d1bxh8uas1mnw7.cloudfront.net/assets/embed.js'></script> <script> jQuery(function(){ load(); }); function load() { jQuery.ajax({ url: '/json/altmetric/get', data: 'hdl=2014.oak/104146', type: 'get', dataType: 'json', success: function(data) { console.log('altmetric : ' + data.response_message); if (data.response_code == 200) { var html = ""; html += "<h4 class='widget_title'>Altmetric</h4>"; html += "<div class='altmetric_area'>"; html += "<div data-badge-details='right' data-condensed='true' data-badge-type='donut' data-"+ data.identifier + "='" + data.identifier +"' data-hide-no-mentions='true' class='altmetric-embed' id='altmetric-embed'></div>"; html += "</div>"; jQuery("#altmetric_donut").html(html); _altmetric_embed_init(); } }, error: function(err) { console.log(err); } }); } </script> <script type="text/javascript" src="https://www.google.com/jsapi"></script> <script type="text/javascript"> google.load("visualization", "1", {packages:["corechart"]}); google.setOnLoadCallback(drawChart); function drawChart() { var data = google.visualization.arrayToDataTable([ ['Type', 'Count', {role:'style'}], ['View', 719, '#DC3912'], ['Download', 0, ''] ]); var options = { width: '95%', height: 280, title: 'Item View & Download Count', legend: { position: "none" }, series: { 0: { axis: 'View' }, // Bind series 0 to an axis named 'distance'. 1: { axis: 'Download' } // Bind series 1 to an axis named 'brightness'. }, axes: { y: { distance: {label: 'Count'} // Left y-axis. } }, bar: { groupWidth: '30' } }; var chart = new google.visualization.ColumnChart(document.getElementById('item_statistics')); chart.draw(data, options); }; (function($){ $(window).resize(function(){ drawChart(); }); })(jQuery.noConflict()); </script> <div class="w_statistics"> <h4 class="widget_title">Views & Downloads</h4> <div class="item_statistics_area"> <div id="item_statistics" class="chart"></div> </div> </div> </div>  </div> </div> <div class="footer_wrap"> <div class="col_width"> <span class="oak_logo">OAK</span> <div class="footer_address"> <div class="footer_link"> <a href="https://www.postech.ac.kr/privacy-policy" target="_blank">개인정보처리방침</a> <a href="https://www.postech.ac.kr/eng/privacy-policy" target="_blank">Personal Information Protection Policy</a> </div> <p><a href="mailto:library@postech.ac.kr" class="t_mail">library@postech.ac.kr</a> <em class="t_phone">Tel: 054-279-2548</em></p> <p>Copyrights © by 2017 Pohang University of Science ad Technology All right reserved.</p> </div> </div> </div> </div> <div class="sb-slidebar sb-right"> <div class="right_nav_box"> <h3>Browse</h3> <ul> <li><a href="/community-list">Communities & Collections</a></li> <li><a href="/browse-researcher">Researcher</a></li> <li><a href="/browse?type=title">Title</a></li> </ul> <div class="left_quick_link"> <a href="/password-login">Login</a> <a href="http://library.postech.ac.kr">Library</a> <a href="#">Help</a> </div> </div> </div>  <script src="/js/slidebars.js"></script> <script> (function($) { $(function() { $("#slide_wrap").owlCarousel({ autoPlay : 3000, navigation : true, slideSpeed : 300, paginationSpeed : 400, singleItem : true }); $.slidebars(); }); })(jQuery.noConflict()); </script> </body> </html>

CINXE.COM

OASIS Repository@POSTECHLIBRARY: 하이 패스 필터와 딥러닝을 이용한 협동로봇의 충돌감지