<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Publishing DTD v1.3 20210610//EN" "JATS-journalpublishing1-3.dtd">
<article article-type="research-article" dtd-version="1.3" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xml:lang="ru"><front><journal-meta><journal-id journal-id-type="publisher-id">bsuir</journal-id><journal-title-group><journal-title xml:lang="ru">Доклады БГУИР</journal-title><trans-title-group xml:lang="en"><trans-title>Doklady BGUIR</trans-title></trans-title-group></journal-title-group><issn pub-type="ppub">1729-7648</issn><issn pub-type="epub">2708-0382</issn><publisher><publisher-name>БГУИР</publisher-name></publisher></journal-meta><article-meta><article-id pub-id-type="doi">10.35596/1729-7648-2026-24-3-69-76</article-id><article-id custom-type="elpub" pub-id-type="custom">bsuir-4377</article-id><article-categories><subj-group subj-group-type="heading"><subject>Research Article</subject></subj-group><subj-group subj-group-type="section-heading" xml:lang="ru"><subject>Статьи</subject></subj-group></article-categories><title-group><article-title>Физически информированная нейронная сеть для решения уравнения конвективной диффузии в природных дисперсных средах</article-title><trans-title-group xml:lang="en"><trans-title>A Physically Informed Neural Network for Solving the Convective Diffusion Equation in Natural Dispersed Media</trans-title></trans-title-group></title-group><contrib-group><contrib contrib-type="author" corresp="yes"><name-alternatives><name name-style="eastern" xml:lang="ru"><surname>Николаенко</surname><given-names>Е. А.</given-names></name><name name-style="western" xml:lang="en"><surname>Nikolaenko</surname><given-names>E.</given-names></name></name-alternatives><bio xml:lang="ru"><p>Николенко Екатерина Анатольевна, асп. каф. информационных технологий в экологии и медицине</p><p>220037, Минск, ул. Долгобродская, 23/1</p><p>Тел.: +375 29 580-84-79</p></bio><bio xml:lang="en"><p>Nikolaenko E., Postgraduate of Information Technologies in Ecology and Medicine Department</p><p>220037, Minsk, Dolgobrodskaya St., 23/1</p><p>Tel.: +375 29 580-84-79</p></bio><email xlink:type="simple">nikolaenko@iseu.by</email><xref ref-type="aff" rid="aff-1"/></contrib><contrib contrib-type="author" corresp="yes"><name-alternatives><name name-style="eastern" xml:lang="ru"><surname>Шалькевич</surname><given-names>П. К.</given-names></name><name name-style="western" xml:lang="en"><surname>Shalkevich</surname><given-names>P.</given-names></name></name-alternatives><bio xml:lang="ru"><p>Шалькевич П. К., канд. тех. наук., доц. каф. информационных технологий в экологии и медицине</p><p>Минск</p></bio><bio xml:lang="en"><p>Shalkevich P., Cand. Sci. (Tech.), Associate Professor of the Department of Information Technologies in Ecology and Medicine</p><p>Minsk</p></bio><xref ref-type="aff" rid="aff-1"/></contrib></contrib-group><aff-alternatives id="aff-1"><aff xml:lang="ru"><institution>Международный государственный экологический институт имени А. Д. Сахарова Белорусского государственного университета</institution></aff><aff xml:lang="en"><institution>International Sakharov Environmental Institute of Belarusian State University</institution></aff></aff-alternatives><pub-date pub-type="collection"><year>2026</year></pub-date><pub-date pub-type="epub"><day>29</day><month>06</month><year>2026</year></pub-date><volume>24</volume><issue>3</issue><fpage>69</fpage><lpage>76</lpage><permissions><copyright-statement>Copyright &amp;#x00A9; Николаенко Е.А., Шалькевич П.К., 2026</copyright-statement><copyright-year>2026</copyright-year><copyright-holder xml:lang="ru">Николаенко Е.А., Шалькевич П.К.</copyright-holder><copyright-holder xml:lang="en">Nikolaenko E., Shalkevich P.</copyright-holder><license xml:lang="ru" license-type="creative-commons-attribution" xlink:href="https://creativecommons.org/licenses/by/4.0/" xlink:type="simple"><license-p>Данная работа распространяется под лицензией Creative Commons Attribution 4.0.</license-p></license><license xml:lang="en" license-type="creative-commons-attribution" xlink:href="https://creativecommons.org/licenses/by/4.0/" xlink:type="simple"><license-p>This work is licensed under a Creative Commons Attribution 4.0 License.</license-p></license></permissions><self-uri xlink:href="https://doklady.bsuir.by/jour/article/view/4377">https://doklady.bsuir.by/jour/article/view/4377</self-uri><abstract><p>Рассмотрено применение физически информированных нейронных сетей для моделирования процессов миграции загрязняющих веществ в природных дисперсных средах. В качестве базовой математической модели использовалось одномерное уравнение конвективной диффузии, описывающее перенос вещества под действием адвективных и диффузионных механизмов. Предложена архитектура нейронной сети на основе многослойного персептрона, позволяющая аппроксимировать решение в непрерывной пространственно-временной области без необходимости построения расчетной сетки. Обучение модели осуществлялось путем минимизации функции потерь, включающей невязку дифференциального уравнения, а также отклонения от начальных и граничных условий. Для обучения использовались коллокационные точки, генерируемые внутри расчетной области. Проведенные вычислительные эксперименты показали, что разработанная модель корректно воспроизводит основные физические закономерности процесса переноса, включая смещение максимума концентрации вследствие адвекции и его сглаживание за счет диффузии. Показано, что применение физически информированных нейронных сетей обеспечивает получение гладкого, устойчивого и физически согласованного решения даже при ограниченном объеме исходной информации. Отмечены преимущества метода, связанные с отсутствием необходимости в обучающих данных и возможностью работы в областях сложной геометрии.</p></abstract><trans-abstract xml:lang="en"><p>This paper examines the application of physically informed neural networks to modeling pollutant migration processes in natural dispersed media. A one-dimensional convective diffusion equation describing substance transport under the influence of advective and diffusion mechanisms was used as the basic mathematical model. A neural network architecture based on a multilayer perceptron is proposed, enabling the approximation of the solution in a continuous spatiotemporal domain without the need to construct a computational grid. The model was trained by minimizing the loss function, which includes the residual of the differential equation, as well as deviations from the initial and boundary conditions. Collocation points generated within the computational domain were used for training. Computational experiments demonstrated that the developed model correctly reproduces the fundamental physical laws of the transport process, including the shift of the maximum concentration due to advection and its smoothing due to diffusion. It is demonstrated that the use of physically informed neural networks ensures a smooth, stable, and physically consistent solution even with a limited amount of initial information. The advantages of the method are noted, which include the absence of the need for training data and the ability to work in areas of complex geometry.</p></trans-abstract><kwd-group xml:lang="ru"><kwd>нейронные сети</kwd><kwd>физически информированные нейронные сети</kwd><kwd>конвективная диффузия</kwd><kwd>функция потерь</kwd><kwd>моделирование загрязняющих веществ</kwd></kwd-group><kwd-group xml:lang="en"><kwd>neural networks</kwd><kwd>physics-informed neural networks</kwd><kwd>convection–diffusion</kwd><kwd>loss function</kwd><kwd>pollutant transport modeling</kwd></kwd-group></article-meta></front><back><ref-list><title>References</title><ref id="cit1"><label>1</label><citation-alternatives><mixed-citation xml:lang="ru">Кундас, С. П. Разработка нейронных сетей для прогнозирования миграции химических веществ в почве и алгоритмов их обучения / С. П. Кундас, В. И. Коваленко, О. С. Хилько // Вестник БНТУ. 2010. № 2. С. 32–38.</mixed-citation><mixed-citation xml:lang="en">Kundas S. P., Kovalenko V. I., Khilko O. S. (2010) Development of Neural Networks for Predicting Migration of Chemical Substances in Soil and Algorithms for Their Training. Bulletin of BNTU. (2), 32–38 (in Russian).</mixed-citation></citation-alternatives></ref><ref id="cit2"><label>2</label><citation-alternatives><mixed-citation xml:lang="ru">Компьютерное моделирование миграции загрязняющих веществ в природных дисперсных средах / С. П. Кундас [и др.]. Минск: Междунар. гос. экологич. ин-т им. А. Д. Сахарова, 2011.</mixed-citation><mixed-citation xml:lang="en">Kundas S. P., Gishkelyuk I. A., Kovalenko V. I., Khilko O. S. (2011) Computer Simulation of the Migration of Pollutants in Natural Dispersed Media. Minsk, International Sakharov Environmental University (in Russian).</mixed-citation></citation-alternatives></ref><ref id="cit3"><label>3</label><citation-alternatives><mixed-citation xml:lang="ru">Шалькевич, П. К. Компьютерное прогнозирование пространственного распределения концентрации Cs-137 в почве / П. К. Шалькевич // Доклады Национальной академии наук Беларуси. 2021. Т. 65, № 2. С. 139–145. https://doi.org/10.29235/1561-8323-2021-65-2-139-145.</mixed-citation><mixed-citation xml:lang="en">Shalkevich P. K. (2021) Computer Prediction of the Spatial Distribution of the Cs-137 Concentration in Soil. Doklady of the National Academy of Sciences of Belarus. 65 (2), 139–145. http://dx.doi.org/10.29235/1561-8323-2021-65-2-139-145 (in Russian).</mixed-citation></citation-alternatives></ref><ref id="cit4"><label>4</label><citation-alternatives><mixed-citation xml:lang="ru">Raissi, М. Physics-Informed Neural Networks: A Deep Learning Framework for Solving Forward and Inverse Problems Involving Nonlinear Partial Diﬀerential Equations / M. Raissi, P. Perdikaris, G. E. Karniadakis // Journal of Computational Physics. 2019. Vol. 378. P. 686–707.</mixed-citation><mixed-citation xml:lang="en">Raissi M., Perdikaris P., Karniadakis G. E. (2019) Physics-Informed Neural Networks: A Deep Learning Framework for Solving Forward and Inverse Problems Involving Nonlinear Partial Diﬀerential Equations. Journal of Computational Physics. Elsevier. 378, 686–707.</mixed-citation></citation-alternatives></ref><ref id="cit5"><label>5</label><citation-alternatives><mixed-citation xml:lang="ru">Physics-Informed Neural Networks for PDE Problems: A Comprehensive Review / K. Luo [et al.] // Artiﬁcial Intelligence Review. 2025. Vol. 58. Article 323. http://dx.doi.org/10.1007/s10462-025-11322-7.</mixed-citation><mixed-citation xml:lang="en">Luo K., Zhao J., Wang Y., Li J., Wen J., Liang J., et al. (2025) Physics-Informed Neural Networks for PDE Problems: A Comprehensive Review. Artiﬁcial Intelligence Review. Springer. 58, Article 323. http://dx.doi.org/10.1007/s10462-025-11322-7.</mixed-citation></citation-alternatives></ref><ref id="cit6"><label>6</label><citation-alternatives><mixed-citation xml:lang="ru">Pre-Trained Physics-Informed Neural Networks for Analysis of Contaminant Transport in Soils / Z.-W. Ke [et al.] // Computers and Geotechnics. 2025. Vol. 180, No 11. http://dx.doi.org/10.1016/j.compgeo.2025.107055.</mixed-citation><mixed-citation xml:lang="en">Ke Z.-W., Wei S.-J., Yao S.-Y., Chen S., Chen Y.-M., Li Y.-C. (2025) Pre-Trained Physics-Informed Neural Networks for Analysis of Contaminant Transport in Soils. Computers and Geotechnics. Elsevier. 180 (11). http://dx.doi.org/10.1016/j.compgeo.2025.107055.</mixed-citation></citation-alternatives></ref><ref id="cit7"><label>7</label><citation-alternatives><mixed-citation xml:lang="ru">Berardi, М. Inverse Physics-Informed Neural Networks for Transport Models in Porous Materials / M. Berardi, F. V. Difonzo, M. Icardi // Computer Methods in Applied Mechanics and Engineering. 2025. Vol. 435. Article 117628. http://dx.doi.org/10.1016/j.cma.2024.117628.</mixed-citation><mixed-citation xml:lang="en">Berardi M., Difonzo F. V., Icardi M. (2025) Inverse Physics-Informed Neural Networks for Transport Models in Porous Materials. Computer Methods in Applied Mechanics and Engineering. Elsevier. 435, Article 117628. http://dx.doi.org/10.1016/j.cma.2024.117628.</mixed-citation></citation-alternatives></ref><ref id="cit8"><label>8</label><citation-alternatives><mixed-citation xml:lang="ru">Yin, R. Prediction of Soil Pollution Spatial Distribution Using Physics-Informed Neural Network Based on Spatial Probability Distribution of Pollutants / R. Yin, L. Wang, T. Hu // SSRN Electronic Journal. 2024. https://doi.org/10.2139/ssrn.4918378.</mixed-citation><mixed-citation xml:lang="en">Yin R., Wang L., Hu T. (2024) Prediction of Soil Pollution Spatial Distribution Using Physics-Informed Neural Network Based on Spatial Probability Distribution of Pollutants. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.4918378.</mixed-citation></citation-alternatives></ref><ref id="cit9"><label>9</label><citation-alternatives><mixed-citation xml:lang="ru">Урвачев, П. М. Передовые методы оптимизации работы с нейросетями на современных архитектурах / П. М. Урвачев, В. А. Ковтун // Современные инновации, системы и технологии. 2024. Т. 4, № 4. С. 199–212. https://doi.org/10.47813/2782-2818-2024-4-4-0199-0212.</mixed-citation><mixed-citation xml:lang="en">Urvachev P. M., Kovtun V. A. (2024) Advanced Methods for Optimizing Work with Neural Networks on Modern Architectures. Modern Innovations, Systems and Technologies. 4 (4), 199–212. http://dx.doi.org/10.47813/2782-2818-2024-4-4-0199-0212 (in Russian).</mixed-citation></citation-alternatives></ref></ref-list><fn-group><fn fn-type="conflict"><p>The authors declare that there are no conflicts of interest present.</p></fn></fn-group></back></article>
