Powered by OpenAIRE graph
Found an issue? Give us feedback
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/ ZENODOarrow_drop_down
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
ZENODO
Article . 2023
License: CC BY
Data sources: ZENODO
ZENODO
Article . 2023
License: CC BY
Data sources: Datacite
addClaim

Алгоритмы редукции сложности математических моделей для обеспечения работы цифрового двойника в режиме мягкого реального времени

Algorithms for reducing the complexity of mathematical models to ensure the operation of a digital twin in soft real time

Алгоритмы редукции сложности математических моделей для обеспечения работы цифрового двойника в режиме мягкого реального времени

Abstract

Статья посвящена анализу алгоритмов редукции сложности математических моделей, применяемых для обеспечения работы цифровых двойников в режиме мягкого реального времени. Показано, что традиционные высокоразмерные FEM/CFD-модели, ориентированные на офлайн-расчёты, не удовлетворяют требованиям по латентности и устойчивости при непрерывной обработке потоков телеметрии. На основе обзора современных публикаций систематизированы архитектурные решения построения цифровых двойников, выделены ключевые классы методов MOR и ROM, включая проекционные, POD-подходы и data-driven/гибридные схемы с использованием нейросетевых аппроксиматоров. Приведены иллюстративные примеры из энергетики и ветроэнергетики, демонстрирующие многократное сокращение вычислительных затрат и измеримый эксплуатационный эффект (рост КПД, уменьшение времени пуска, повышение устойчивости режимов). Отдельное внимание уделено ограничениям и рискам: потере точности за пределами области обучающих данных, потенциальной утрате устойчивости и необходимости постоянной валидации ROM. Сформулированы перспективы развития, связанные с гибридными моделями, нейросетевыми редукторами и онлайновой адаптацией параметров цифрового двойника.

The article is devoted to the analysis of algorithms for reducing the complexity of mathematical models used to ensure the operation of digital twins in soft real time. It is shown that traditional high-dimensional FEM/CFD models focused on offline calculations do not meet the requirements for latency and stability during continuous processing of telemetry streams. Based on a review of modern publications, architectural solutions for building digital twins are systematized, and key classes of MOR and ROM methods are identified, including projection, POD approaches, and data-driven/hybrid schemes using neural network approximators. Illustrative examples from the energy and wind power industries are presented, demonstrating a multiple reduction in computing costs and a measurable operational effect (increased efficiency, reduced start-up time, increased stability of modes). Special attention is paid to the limitations and risks: loss of accuracy beyond the training data domain, potential loss of stability, and the need for constant ROM validation. Development prospects related to hybrid models, neural network reducers, and online adaptation of digital twin parameters are formulated.

Keywords

режим мягкого реального времени, MOR, ROM, энергетика, редукция порядка модели, цифровой двойник, киберфизическая система, машинное обучение

  • BIP!
    Impact byBIP!
    selected citations
    These citations are derived from selected sources.
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    0
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Average
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Average
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Average
Powered by OpenAIRE graph
Found an issue? Give us feedback
selected citations
These citations are derived from selected sources.
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
0
Average
Average
Average
Green