基于物理嵌入神经网络的叶栅损失模型

Yunyang Feng; Xizhen Song; Wei Yuan; Hanan Lu

doi:10.13224/j.cnki.jasp.20220750

基于物理嵌入神经网络的叶栅损失模型

Translated title of the contribution: Physics-informed neural networks based cascade loss model

Yunyang Feng
, Xizhen Song
, Wei Yuan^*
, Hanan Lu

^*Corresponding author for this work

School of Energy and Power Engineering

Beihang University

Research output: Contribution to journal › Article › peer-review

5 Scopus citations

Abstract

It is difficult to modify and broaden the scope of application for empirical models because of its inadequate ability to fit strong nonlinear function relationship. In order to solve these problems, a physics-informed deep learning cascade loss model of embedding the pressure distribution of cascade into neural networks was proposed. The loss prediction error decreased by 22.3% compared with empirical model for end-to-end neural networks and 37.9% compared with physics-informed model.

Translated title of the contribution	Physics-informed neural networks based cascade loss model
Original language	Chinese (Traditional)
Pages (from-to)	845-855
Number of pages	11
Journal	Hangkong Dongli Xuebao/Journal of Aerospace Power
Volume	38
Issue number	7
DOIs	https://doi.org/10.13224/j.cnki.jasp.20220750
State	Published - Jul 2023

Access to Document

10.13224/j.cnki.jasp.20220750

Cite this

@article{f0715cfdb4fb49bca701bfbbfd66871f,

title = "基于物理嵌入神经网络的叶栅损失模型",

abstract = "It is difficult to modify and broaden the scope of application for empirical models because of its inadequate ability to fit strong nonlinear function relationship. In order to solve these problems, a physics-informed deep learning cascade loss model of embedding the pressure distribution of cascade into neural networks was proposed. The loss prediction error decreased by 22.3\% compared with empirical model for end-to-end neural networks and 37.9\% compared with physics-informed model.",

keywords = "deep learning, loss prediction, neural networks, physics-informed, surrogate model",

author = "Yunyang Feng and Xizhen Song and Wei Yuan and Hanan Lu",

year = "2023",

month = jul,

doi = "10.13224/j.cnki.jasp.20220750",

language = "繁体中文",

volume = "38",

pages = "845--855",

journal = "Hangkong Dongli Xuebao/Journal of Aerospace Power",

issn = "1000-8055",

publisher = "BUAA Press",

number = "7",

}

TY - JOUR

T1 - 基于物理嵌入神经网络的叶栅损失模型

AU - Feng, Yunyang

AU - Song, Xizhen

AU - Yuan, Wei

AU - Lu, Hanan

PY - 2023/7

Y1 - 2023/7

N2 - It is difficult to modify and broaden the scope of application for empirical models because of its inadequate ability to fit strong nonlinear function relationship. In order to solve these problems, a physics-informed deep learning cascade loss model of embedding the pressure distribution of cascade into neural networks was proposed. The loss prediction error decreased by 22.3% compared with empirical model for end-to-end neural networks and 37.9% compared with physics-informed model.

AB - It is difficult to modify and broaden the scope of application for empirical models because of its inadequate ability to fit strong nonlinear function relationship. In order to solve these problems, a physics-informed deep learning cascade loss model of embedding the pressure distribution of cascade into neural networks was proposed. The loss prediction error decreased by 22.3% compared with empirical model for end-to-end neural networks and 37.9% compared with physics-informed model.

KW - deep learning

KW - loss prediction

KW - neural networks

KW - physics-informed

KW - surrogate model

UR - https://www.scopus.com/pages/publications/85164288092

U2 - 10.13224/j.cnki.jasp.20220750

DO - 10.13224/j.cnki.jasp.20220750

M3 - 文章

AN - SCOPUS:85164288092

SN - 1000-8055

VL - 38

SP - 845

EP - 855

JO - Hangkong Dongli Xuebao/Journal of Aerospace Power

JF - Hangkong Dongli Xuebao/Journal of Aerospace Power

IS - 7

ER -

基于物理嵌入神经网络的叶栅损失模型

Abstract

Access to Document

Other files and links

Fingerprint

Cite this