Changchun Deng , Tian Qiu , Peng Liu , Shuiting Ding , Xiang Luo
{"title":"利用壁温特征正则化 BP 神经网络降低旋转盘腔中二维逆传热问题的拟合不良性","authors":"Changchun Deng , Tian Qiu , Peng Liu , Shuiting Ding , Xiang Luo","doi":"10.1016/j.ijthermalsci.2024.109145","DOIUrl":null,"url":null,"abstract":"<div><p>In the two-dimensional heat transfer experiments of aero-engine rotating disk cavities, the inverse heat transfer problem method can be used to obtain the wall heat flux numerically, which uses the two-dimensional measured wall temperature to solve the rotating disk heat conduction equation. A back propagation (BP) neural network data approximation method is proposed to reduce the ill-posedness of the two-dimensional inverse heat transfer problems in rotating disk cavities in this paper. The priori knowledge of wall temperature characteristics expressed by two-dimensional wall temperature first-order radial partial derivative distribution is used for BP neural networks’ regularization. The distribution characteristics of the wall temperature first-order radial partial derivative in a typical preswirl rotating disk cavity were investigated by the flow-thermal coupling numerical simulation. Based on these characteristics, the BP neural network construction and training method with uncertain regularization coefficient is adopted. The numerical experiment results show that compared with the traditional polynomial fitting methods, the BP neural network approximation methods in this paper show significant advantages in data processing performance and stability; The fluctuation amplitude of the wall heat flux relative error on the disk surface can be reduced by 1–3 orders of magnitude, reducing the relative error of wall heat flux in most areas of the disk to within 20 % of the original value; The maximum wall heat flux relative error suppression area where |<em>δq</em><sub><em>r</em>,cal</sub>/<em>δq</em><sub><em>r</em>,mea</sub> × 100 %| < 100 % of BP neural network approximation method can reach 1.93 times that of the traditional fitting method, and 3.18 times for the area where |<em>δq</em><sub><em>r</em>,cal</sub>/<em>δq</em><sub><em>r</em>,mea</sub> × 100 %| < 30 % in the current study.</p></div>","PeriodicalId":341,"journal":{"name":"International Journal of Thermal Sciences","volume":null,"pages":null},"PeriodicalIF":4.9000,"publicationDate":"2024-05-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"BP neural network regularized by wall temperature characteristics to reduce the ill-posedness of two-dimensional inverse heat transfer problems in rotating disk cavities\",\"authors\":\"Changchun Deng , Tian Qiu , Peng Liu , Shuiting Ding , Xiang Luo\",\"doi\":\"10.1016/j.ijthermalsci.2024.109145\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><p>In the two-dimensional heat transfer experiments of aero-engine rotating disk cavities, the inverse heat transfer problem method can be used to obtain the wall heat flux numerically, which uses the two-dimensional measured wall temperature to solve the rotating disk heat conduction equation. A back propagation (BP) neural network data approximation method is proposed to reduce the ill-posedness of the two-dimensional inverse heat transfer problems in rotating disk cavities in this paper. The priori knowledge of wall temperature characteristics expressed by two-dimensional wall temperature first-order radial partial derivative distribution is used for BP neural networks’ regularization. The distribution characteristics of the wall temperature first-order radial partial derivative in a typical preswirl rotating disk cavity were investigated by the flow-thermal coupling numerical simulation. Based on these characteristics, the BP neural network construction and training method with uncertain regularization coefficient is adopted. The numerical experiment results show that compared with the traditional polynomial fitting methods, the BP neural network approximation methods in this paper show significant advantages in data processing performance and stability; The fluctuation amplitude of the wall heat flux relative error on the disk surface can be reduced by 1–3 orders of magnitude, reducing the relative error of wall heat flux in most areas of the disk to within 20 % of the original value; The maximum wall heat flux relative error suppression area where |<em>δq</em><sub><em>r</em>,cal</sub>/<em>δq</em><sub><em>r</em>,mea</sub> × 100 %| < 100 % of BP neural network approximation method can reach 1.93 times that of the traditional fitting method, and 3.18 times for the area where |<em>δq</em><sub><em>r</em>,cal</sub>/<em>δq</em><sub><em>r</em>,mea</sub> × 100 %| < 30 % in the current study.</p></div>\",\"PeriodicalId\":341,\"journal\":{\"name\":\"International Journal of Thermal Sciences\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":4.9000,\"publicationDate\":\"2024-05-28\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"International Journal of Thermal Sciences\",\"FirstCategoryId\":\"5\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S1290072924002679\",\"RegionNum\":2,\"RegionCategory\":\"工程技术\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"ENGINEERING, MECHANICAL\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"International Journal of Thermal Sciences","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S1290072924002679","RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, MECHANICAL","Score":null,"Total":0}
引用次数: 0
摘要
在航空发动机旋转盘空腔的二维传热实验中,可采用逆传热问题方法数值求得壁面热通量,该方法利用二维实测壁面温度求解旋转盘热传导方程。本文提出了一种反向传播(BP)神经网络数据逼近方法,以降低旋转盘腔中二维反向传热问题的非拟合性。通过二维壁温一阶径向偏导数分布表示的壁温特征的先验知识用于 BP 神经网络的正则化。通过流热耦合数值模拟研究了典型预旋流旋转盘腔中壁温一阶径向偏导数的分布特征。根据这些特征,采用不确定正则化系数的 BP 神经网络构建和训练方法。数值实验结果表明,与传统的多项式拟合方法相比,本文的 BP 神经网络逼近方法在数据处理性能和稳定性方面具有显著优势;盘面壁面热通量相对误差的波动幅度可降低 1-3 个数量级,将盘面大部分区域的壁面热通量相对误差降低到原始值的 20% 以内;BP 神经网络逼近方法的 |/ × 100 %| < 100 % 的最大壁面热通量相对误差抑制区域可达传统拟合方法的 1.93倍,在本次研究中,|/ × 100 %| < 30 %的区域是传统拟合方法的3.18倍。
BP neural network regularized by wall temperature characteristics to reduce the ill-posedness of two-dimensional inverse heat transfer problems in rotating disk cavities
In the two-dimensional heat transfer experiments of aero-engine rotating disk cavities, the inverse heat transfer problem method can be used to obtain the wall heat flux numerically, which uses the two-dimensional measured wall temperature to solve the rotating disk heat conduction equation. A back propagation (BP) neural network data approximation method is proposed to reduce the ill-posedness of the two-dimensional inverse heat transfer problems in rotating disk cavities in this paper. The priori knowledge of wall temperature characteristics expressed by two-dimensional wall temperature first-order radial partial derivative distribution is used for BP neural networks’ regularization. The distribution characteristics of the wall temperature first-order radial partial derivative in a typical preswirl rotating disk cavity were investigated by the flow-thermal coupling numerical simulation. Based on these characteristics, the BP neural network construction and training method with uncertain regularization coefficient is adopted. The numerical experiment results show that compared with the traditional polynomial fitting methods, the BP neural network approximation methods in this paper show significant advantages in data processing performance and stability; The fluctuation amplitude of the wall heat flux relative error on the disk surface can be reduced by 1–3 orders of magnitude, reducing the relative error of wall heat flux in most areas of the disk to within 20 % of the original value; The maximum wall heat flux relative error suppression area where |δqr,cal/δqr,mea × 100 %| < 100 % of BP neural network approximation method can reach 1.93 times that of the traditional fitting method, and 3.18 times for the area where |δqr,cal/δqr,mea × 100 %| < 30 % in the current study.
期刊介绍:
The International Journal of Thermal Sciences is a journal devoted to the publication of fundamental studies on the physics of transfer processes in general, with an emphasis on thermal aspects and also applied research on various processes, energy systems and the environment. Articles are published in English and French, and are subject to peer review.
The fundamental subjects considered within the scope of the journal are:
* Heat and relevant mass transfer at all scales (nano, micro and macro) and in all types of material (heterogeneous, composites, biological,...) and fluid flow
* Forced, natural or mixed convection in reactive or non-reactive media
* Single or multi–phase fluid flow with or without phase change
* Near–and far–field radiative heat transfer
* Combined modes of heat transfer in complex systems (for example, plasmas, biological, geological,...)
* Multiscale modelling
The applied research topics include:
* Heat exchangers, heat pipes, cooling processes
* Transport phenomena taking place in industrial processes (chemical, food and agricultural, metallurgical, space and aeronautical, automobile industries)
* Nano–and micro–technology for energy, space, biosystems and devices
* Heat transport analysis in advanced systems
* Impact of energy–related processes on environment, and emerging energy systems
The study of thermophysical properties of materials and fluids, thermal measurement techniques, inverse methods, and the developments of experimental methods are within the scope of the International Journal of Thermal Sciences which also covers the modelling, and numerical methods applied to thermal transfer.