Spaces:
Running
Running
File size: 13,854 Bytes
00e479e fd6cfa6 b440e64 fd6cfa6 b440e64 fd6cfa6 b440e64 fd6cfa6 b440e64 fd6cfa6 fe8c37a fd6cfa6 00e479e | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 | <!doctype html>
<html lang="en" data-magnet-static="true">
<head>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1">
<meta name="description" content="Princeton MagNet data, models, training resources, and prediction tools for power magnetics.">
<title>Princeton MagNet</title>
<link rel="stylesheet" href="./assets/styles.css">
</head>
<body>
<a class="skip-link" href="#main">Skip to content</a>
<header class="site-header">
<div class="header-inner">
<a class="brand" href="#/" aria-label="Princeton MagNet home">
<span class="brand-mark" aria-hidden="true">M</span>
<span><strong>MagNet</strong><small>Princeton University</small></span>
</a>
<nav aria-label="Primary navigation">
<a href="#/data" data-route="data">Data</a>
<a href="#/models" data-route="models">Models</a>
<a href="#/predict" data-route="predict">Predict</a>
<a href="#/documentation" data-route="documentation">Documentation</a>
</nav>
</div>
</header>
<main id="main">
<section class="view home-view" data-view="home" tabindex="-1">
<div class="home-intro">
<p class="eyebrow">Princeton Power Electronics Research Lab</p>
<h1>MagNet data and model hub</h1>
<p class="lede">Open resources for magnetic-material measurements, model development, and core-loss prediction.</p>
</div>
<div class="module-grid" aria-label="MagNet modules">
<a class="module-panel" href="#/data">
<span class="module-number">01</span>
<div><h2>Data</h2><p>Raw steady-state and transient measurements, plus training-ready releases.</p></div>
<div class="module-footer"><span><strong id="material-count">—</strong> materials · <strong id="dataset-count">—</strong> dataset families</span><span>Open data <span aria-hidden="true">→</span></span></div>
</a>
<a class="module-panel" href="#/models">
<span class="module-number">02</span>
<div><h2>Models</h2><p>Available baselines, migrated research models, and validation status.</p></div>
<div class="module-footer"><span><strong id="model-count">—</strong> registered models</span><span>Open models <span aria-hidden="true">→</span></span></div>
</a>
<a class="module-panel" href="#/predict">
<span class="module-number">03</span>
<div><h2>Predict</h2><p>Configure a waveform and estimate steady-state magnetic core loss.</p></div>
<div class="module-footer"><span>Interactive tool</span><span>Open predictor <span aria-hidden="true">→</span></span></div>
</a>
<a class="module-panel" href="#/documentation">
<span class="module-number">04</span>
<div><h2>Documentation</h2><p>Data contracts, training examples, model packaging, and usage guides.</p></div>
<div class="module-footer"><span>Guides and resources</span><span>Open documentation <span aria-hidden="true">→</span></span></div>
</a>
</div>
</section>
<section class="view module-view" data-view="data" tabindex="-1" hidden>
<a class="back-link" href="#/"><span aria-hidden="true">←</span> All modules</a>
<header class="module-header">
<div><p class="eyebrow">Data library</p><h1>Datasets</h1></div>
<p>Find raw measurements and training-ready releases. Published releases link to immutable files on Hugging Face.</p>
</header>
<div class="library-note">
<span class="hf-mark" aria-hidden="true">HF</span>
<div><strong>Hugging Face releases</strong><p>Only verified public releases show download links. Planned releases remain visible with their publishing status.</p></div>
</div>
<div class="library-toolbar">
<label class="library-search" for="dataset-search">Search datasets<input id="dataset-search" type="search" placeholder="Material, task, format, or title" autocomplete="off"></label>
<div class="filters" role="group" aria-label="Dataset filters">
<button class="filter active" data-dataset-filter="all" aria-pressed="true">All</button>
<button class="filter" data-dataset-filter="steady-state-loss" aria-pressed="false">Steady state</button>
<button class="filter" data-dataset-filter="transient-sequence" aria-pressed="false">Transient</button>
<button class="filter" data-dataset-filter="raw" aria-pressed="false">Raw</button>
<button class="filter" data-dataset-filter="processed" aria-pressed="false">Training ready</button>
<button class="filter" data-dataset-filter="published" aria-pressed="false">Downloadable</button>
</div>
<p id="dataset-result-count" class="result-count" aria-live="polite"></p>
</div>
<div id="dataset-grid" class="catalog-list" aria-live="polite"><p class="loading">Loading datasets…</p></div>
</section>
<section class="view module-view" data-view="models" tabindex="-1" hidden>
<a class="back-link" href="#/"><span aria-hidden="true">←</span> All modules</a>
<header class="module-header">
<div><p class="eyebrow">Model library</p><h1>Models</h1></div>
<p>Browse baselines and trained model packages. Published releases link to pinned artifacts and model cards on Hugging Face.</p>
</header>
<div class="library-note">
<span class="hf-mark" aria-hidden="true">HF</span>
<div><strong>Hugging Face releases</strong><p>Downloads are enabled only after the repository, immutable revision, license, and file checksums are verified.</p></div>
</div>
<div class="library-toolbar">
<label class="library-search" for="model-search">Search models<input id="model-search" type="search" placeholder="Material, framework, task, or title" autocomplete="off"></label>
<div class="filters" role="group" aria-label="Model filters">
<button class="filter active" data-model-filter="all" aria-pressed="true">All</button>
<button class="filter" data-model-filter="published" aria-pressed="false">Downloadable</button>
<button class="filter" data-model-filter="inference" aria-pressed="false">Online predictor</button>
<button class="filter" data-model-filter="steady-state-loss" aria-pressed="false">Steady state</button>
<button class="filter" data-model-filter="transient-sequence" aria-pressed="false">Transient</button>
</div>
<p id="model-result-count" class="result-count" aria-live="polite"></p>
</div>
<div id="model-grid" class="catalog-list" aria-live="polite"><p class="loading">Loading models…</p></div>
</section>
<section class="view module-view" data-view="predict" tabindex="-1" hidden>
<a class="back-link" href="#/"><span aria-hidden="true">←</span> All modules</a>
<header class="module-header">
<div><p class="eyebrow">Predict</p><h1>Compare core-loss predictions</h1></div>
<p id="predict-description">Enter one steady-state operating point to see the iGSE baseline and MagNet Transformer estimates together, plus the Transformer's predicted B–H loop. Model limitations and operating-range warnings appear with each result.</p>
</header>
<div class="prediction-layout">
<form id="prediction-form" class="prediction-form">
<label>Material<select id="material" name="material" required></select></label>
<label>Waveform<select id="waveform" name="waveform"><option value="sine">Sinusoidal</option><option value="triangle">Triangular</option></select></label>
<label id="duty-control" hidden>Rising duty ratio <span><output id="duty-output">50%</output></span><input id="duty" type="range" min="0.01" max="0.99" value="0.50" step="0.01" aria-describedby="duty-help"><small id="duty-help">Fraction of one cycle used to ramp from −Bpk to +Bpk.</small></label>
<label>Frequency <span><output id="frequency-output">100</output> kHz</span><input id="frequency" type="range" min="10" max="1000" value="100" step="5"></label>
<label>Peak flux density <span><output id="flux-output">100</output> mT</span><input id="flux" type="range" min="5" max="400" value="100" step="5"></label>
<label>Temperature <span><output id="temperature-output">25</output> °C</span><input id="temperature" type="range" min="0" max="120" value="25" step="5"></label>
<label id="bias-control">DC bias field <span><output id="bias-output">0</output> A/m</span><input id="bias" type="range" min="0" max="75" value="0" step="5"><small>Used by the Transformer; iGSE does not model DC bias or temperature.</small></label>
<button id="predict-button" class="button primary" type="submit">Calculate both models</button>
</form>
<div class="prediction-output" aria-live="polite">
<div class="waveform-chart"><canvas id="waveform-chart" width="760" height="260" aria-label="Input flux-density waveform">Flux-density waveform preview.</canvas></div>
<p id="waveform-summary" class="waveform-summary">Sinusoidal waveform, 100 mT peak.</p>
<div class="prediction-results" aria-label="Predicted volumetric loss density">
<article class="loss-card">
<h2>iGSE baseline</h2>
<strong id="igse-loss-result">—</strong>
<small>Predicted volumetric loss</small>
<p id="igse-prediction-note">Uses waveform, frequency, and material; ignores temperature and DC bias.</p>
</article>
<article class="loss-card">
<h2>MagNet Transformer</h2>
<strong id="neural-loss-result">—</strong>
<small>Predicted volumetric loss</small>
<p id="neural-prediction-note">Experimental legacy model. First run downloads the browser runtime and material model.</p>
</article>
</div>
<section id="bh-loop-section" class="bh-loop-section" aria-labelledby="bh-loop-heading" hidden>
<h2 id="bh-loop-heading">Predicted B–H loop</h2>
<div class="bh-loop-chart">
<svg id="bh-loop-svg" viewBox="0 0 760 360" role="img" aria-labelledby="bh-loop-title bh-loop-description">
<title id="bh-loop-title">Transformer-predicted B–H loop</title>
<desc id="bh-loop-description">Predicted field strength H on the horizontal axis and input flux density B on the vertical axis, in sample order.</desc>
<line class="bh-axis" x1="74" y1="170" x2="728" y2="170"></line>
<line class="bh-axis" x1="401" y1="40" x2="401" y2="300"></line>
<path id="bh-loop-path" d="" fill="none"></path>
<circle id="bh-loop-start" r="4" cx="0" cy="0"></circle>
<text id="bh-h-negative" x="74" y="326" text-anchor="start"></text>
<text id="bh-h-positive" x="728" y="326" text-anchor="end"></text>
<text id="bh-b-positive" x="76" y="34" text-anchor="start"></text>
<text id="bh-b-negative" x="76" y="298" text-anchor="start"></text>
<text class="bh-axis-label" x="401" y="348" text-anchor="middle">H (A/m)</text>
<text class="bh-axis-label" x="30" y="170" transform="rotate(-90 30 170)" text-anchor="middle">B (mT)</text>
</svg>
</div>
<p class="bh-loop-note">128 paired samples, in cycle order. The trace is not artificially closed; Transformer loss follows the original 127-interval H dB integration.</p>
</section>
</div>
</div>
</section>
<section class="view module-view" data-view="documentation" tabindex="-1" hidden>
<a class="back-link" href="#/"><span aria-hidden="true">←</span> All modules</a>
<header class="module-header">
<div><p class="eyebrow">Documentation</p><h1>Use MagNet</h1></div>
<p>Follow the shared data and model contracts so experiments remain reproducible and published artifacts remain traceable.</p>
</header>
<ol class="learning-list">
<li><strong>Review the measurements</strong><span>Understand equipment, geometry, units, calibration, and acquisition metadata.</span></li>
<li><strong>Load a dataset</strong><span>Use stable sample identifiers and published train, validation, and test splits.</span></li>
<li><strong>Train a baseline</strong><span>Run the included steady-state or transient example with a recorded configuration.</span></li>
<li><strong>Package a model</strong><span>Use the common inference interface and document the model's operating limits.</span></li>
</ol>
<div class="resource-links">
<a href="https://huggingface.co/datasets/minjiechen/MagNet" target="_blank" rel="noopener" data-api-reference><strong>Dataset guide</strong><span>Published data and download instructions ↗</span></a>
<a href="https://github.com/PrincetonUniversity/magnet"><strong>Source repository</strong><span>Code, examples, and issue tracking <span aria-hidden="true">→</span></span></a>
</div>
</section>
</main>
<footer>
<div><strong>Princeton MagNet</strong><span>Power Electronics Research Lab</span></div>
<div><a href="https://github.com/PrincetonUniversity/magnet">GitHub</a><a href="https://www.princeton.edu/~minjie/magnet.html">Original project</a><a href="#/documentation" data-api-footer>Documentation</a></div>
</footer>
<script src="./assets/static-predictor.js" defer></script>
<script src="./assets/neural-predictor.js" defer></script>
<script src="./assets/neural-domain.js" defer></script>
<script src="./assets/app.js" defer></script>
</body>
</html>
|