Fast Prediction Model for Gas Turbine Metal Temperature in Digital Twin Applications

LI Jiashan; QUE Xiaobin; YAN Xuehui; WANG Xiaobo; WANG Xu

doi:10.19805/j.cnki.jcspe.2026.250660

English Version Journal Login

您当前的位置：

首页 >

文章列表页 >

Fast Prediction Model for Gas Turbine Metal Temperature in Digital Twin Applications

更新时间：2026-10-06

- Fast Prediction Model for Gas Turbine Metal Temperature in Digital Twin Applications
- Chinese Journal of Power Engineering Vol. 46, Issue 3, Pages: 190-200(2026)
- 作者机构：
  
  中国联合重型燃气轮机技术有限公司,北京,100016
- 作者简介：
- 基金信息：
- DOI：10.19805/j.cnki.jcspe.2026.250660
  CLC：
- Online First：10 March 2026，
  
  Published：10 March 2026
- 稿件说明：
移动端阅览
李甲珊,阙晓斌,闫学慧,王晓博,王旭. 面向数字孪生应用的燃气轮机金属温度快速预测模型研究动力工程学报, 2026, 46(3): 190-200 https://doi.

org/10.19805/j.cnki.jcspe.2026.250660
李甲珊,阙晓斌,闫学慧,王晓博,王旭. 面向数字孪生应用的燃气轮机金属温度快速预测模型研究动力工程学报, 2026, 46(3): 190-200 https://doi. DOI： 10.19805/j.cnki.jcspe.2026.250660.

org/10.19805/j.cnki.jcspe.2026.250660 DOI：

摘要

针对燃气轮机数字孪生应用中的金属温度实时预测需求

提出了一种基于循环神经网络(RNN)的燃气轮机金属温度快速预测方法。传统基于运行曲线工况驱动的整机温度场物理模型存在复杂度高、计算耗时长等缺陷

难以满足燃气轮机试验、运行过程中实时预测与故障预警等应用需求。为解决这一问题

基于物理模型仿真数据集和整机试验数据集训练了2组模型

且经优化后2组模型的保真度分别达到99.14%和94.32%。系统分析了神经网络超参数对模型性能的影响规律

并提出了有效的模型调优和降低过拟合的方法。最后通过测试数据集验证了模型泛化能力

并基于功能模型接口(FMI)标准实现了燃气轮机金属温度快速预测模型在自主F级重型燃气轮机试验样机中的部署应用。研究成果为燃气轮机数字孪生系统的实时温度预测提供了有效的解决方案。

Abstract

To address the real-time prediction requirements of gas turbine metal temperature in digital twin applications

a fast prediction method of gas turbine metal temperature based on recurrent neural networks (RNN) was proposed. Conventional physics-based whole-engine temperature field models driven by operational curves suffer from high complexity and long computation time

failing to meet the demands of real-time prediction and fault warning in gas turbine testing and operation. Two models were trained using datasets from physics-model simulations and whole-engine experiments

achieving fidelity of 99.14% and 94.32% after optimization. The influence of neural network hyperparameters on model performance was systematically analyzed

and effective tuning methods for model optimization and overfitting reduction were proposed. Model generalization capability was verified through test datasets. Deployment was implemented in a prototype F-class heavy-duty gas turbine using functional mock-up interface (FMI) standards

providing an effective solution for real-time temperature prediction in gas turbine digital twin systems.

关键词

Keywords

references

GRIEVES M W. Product lifecycle management: the new paradigm for enterprises[J]. International Journal of Product Development, 2005, 2(1/2): 71-84, doi: 10.1504/IJPD.2005.006669.

HARTMANN D, HERZ M, PAFFRATH M, et al. 12 model order reduction and digital twins[M] //BENNER P, ALFIO Q, ANDREA M, et al. Model order reduction. Volume 3, Applications. Boston: De Gruyter, 2021, doi: 10.1515/9783110499001-012.

MICHAEL G. Digital twin:manufacturing excellence through virtual factory replication[R]. [S.l.] :White Paper,2014.

辛小鹏, 谭建荣, 刘振宇, 等. 燃气轮机正向设计研究进展[J]. 机械工程学报, 2022, 58(17): 191-205.XIN Xiaopeng, TAN Jianrong, LIU Zhengyu, et al. Research progress on forward design of gas turbine[J]. Journal of Mechanical Engineering, 2022, 58(17): 191-205.

YAN Xin, GUO Xinyu, HE Ningya, et al. Model reduction of structural mechanical response in the time domain[J]. Scientific Reports, 2025, 15(1): 10203, doi: 10.1038/s41598-025-94681-z.

冯俞楷. POD降维算法在传热与流动数值模拟中的应用[D]. 北京: 华北电力大学, 2017.

GAO Wansheng, CHAMOIN L, NEUWEILER I. Non-intrusive reduced basis two-grid method for flow and transport problems in heterogeneous porous media[J]. Journal of Computational and Applied Mathematics, 2025, 457: 116321, doi: 10.1016/j.cam.2024.116321.

MüCKE N T, BOHT S M, OOSTERLEE C W. Reduced order modeling for parameterized time-dependent PDEs using spatially and memory aware deep learning[J]. Journal of Computational Science, 2021, 53: 101408, doi: 10.1016/j.jocs.2021.101408.

杨迪, 段承杰, 丁鹏, 等. 参数化流动传热问题的模型降阶方法研究[J]. 原子能科学技术, 2024, 58(7): 1440-1451.YANG Di, DUAN Chengjie, DING Peng, et al. Study of reduced order model for parameterized flow and heat transfer problems[J]. Atomic Energy Science and Technology, 2024, 58(7): 1440-1451.

SIDERIS I, DUNCAN S, FABBRI M, et al. End-to-end path planning for homogeneous temperature fields in additive manufacturing[J]. Journal of Materials Processing Technology, 2024, 327: 118364, doi: 10.1016/j.jmatprotec.2024.118364.

QUARTERONI A, ROZZA G, MANZONI A. Certified reduced basis approximation for parametrized partial differential equations and applications[J]. Journal of Mathematics in Industry, 2011, 1(1): 3, doi: 10.1186/2190-5983-1-3.

姚涵毅, 丁炯, 许金鑫, 等. 基于Krylov子空间降阶的温控算法研究[J]. 中国测试, 2024, 50(2): 146-154.YAO Hanyi, DING Jiong, XU Jinxin, et al. Research of temperature control algorithm based on Krylov subspace reduction[J]. China Measurement & Test, 2024, 50(2): 146-154.

冯俞楷, 杜小泽, 杨立军. 非稳态导热基于温度梯度的本征正交分解降维方法[J]. 中国科学: 技术科学, 2018, 48(1): 39-47.FENG Yukai, DU Xiaoze, YANG Lijun. Extrapolating POD reduced-order model based on temperature gradient for unsteady heat conduction[J]. Scientia Sinica Technologica, 2018, 48(1): 39-47.

YAN Feng, JIAO Kaituo, NIE Chaofei, et al. Fast prediction of the temperature field surrounding a hot oil pipe using the POD-BP model[J]. Processes, 2023, 11(9): 2666.

PENG Xiaojun, CHEN Zhangdong, ZHANG Aoming, et al. Digital twin temperature field prediction of laser powder bed fusion through proper orthogonal decomposition with radial basis function[J]. Materials Today Communications, 2024, 38: 107883, doi: 10.1016/j.mtcomm.2023.107883.

[CM(22*2] CHEN Fan, YANG M, YAN Wentao. Data-driven prognostic model for temperature field in additive manufacturing based on the high-fidelity thermal-fluid flow simulation[J]. Computer Methods in Applied Mechanics and Engineering, 2022, 392: 114652.

韩国栋, 曹云鹏, 王伟影, 等. 基于LSTM神经网络的燃气轮机排温预测方法[J]. 热能动力工程, 2022, 37(3): 28-34.HAN Guodong, CAO Yunpeng, WANG Weiying, et al. Prediction method of gas turbine exhaust temperature based on LSTM neural network[J]. Journal of Engineering for Thermal Energy and Power, 2022, 37(3): 28-34.

李永华, 陈伟华, 谢英柏. 基于数据驱动的燃气轮机性能监测方法[J]. 动力工程学报, 2024, 44(12): 1845-1853.LI Yonghua, CHEN Weihua, XIE Yingbai. Data-driven gas turbine performance monitoring method[J]. Journal of Chinese Society of Power Engineering, 2024, 44(12): 1845-1853.

杨永, 王瑞锋. 基于循环神经网络的动车组温度数据预测研究[J]. 大连交通大学学报, 2024, 45(3): 53-57.YANG Yong, WANG Ruifeng. Research on temperature data prediction of EMU based on cyclic neural network[J]. Journal of Dalian Jiaotong University, 2024, 45(3): 53-57.

姚逸行, 刘建国. 基于LSTM的轨道结构温度变形预测方法[J]. 铁道建筑技术, 2021(9): 33-37.YAO Yihang, LIU Jianguo. Temperature deformation prediction method of track structure based on LSTM[J]. Railway Construction Technology, 2021(9): 33-37.

高尚鸿, 汪翔宇, 张韦馨, 等. 数据驱动在透平叶片冷却设计中的应用进展[J]. 动力工程学报, 2024, 44(11): 1780-1789.GAO Shanghong, WANG Xiangyu, ZHANG Weixin, et al. Progresses on the application of data-driven methods in turbine blade cooling design[J]. Journal of Chinese Society of Power Engineering, 2024, 44(11): 1780-1789.

刘岩松, 阙晓斌, 吴帅, 等. 基于神经网络的重型燃气轮机压气机叶型性能预测[J]. 热能动力工程, 2025, 40(5): 21-31.LIU Yansong, QUE Xiaobin, WU Shuai, et al. Prediction of heavy-duty gas turbine compressor blade performance based on neural network[J]. Journal of Engineering for Thermal Energy and Power, 2025, 40(5): 21-31.

赵宇, 郑学刚, 吴洪涛. 温度传感器失效检测与恢复研究[J]. 机械科学与技术, 2010, 29(6): 805-808.ZHAO Yu, ZHENG Xuegang, WU Hongtao. Failure detection of temperature sensors and their recovery[J]. Mechanical Science and Technology for Aerospace Engineering, 2010, 29(6): 805-808.

王东. 机器学习导论[M]. 北京: 清华大学出版社, 2021: 90-93.

李国良, 周煊赫. 面向AI的数据管理技术综述[J]. 软件学报, 2021, 32(1): 21-40.LI Guoliang, ZHOU Xuanhe. Survey of data management techniques for artificial intelligence[J]. Journal of Software, 2021, 32(1): 21-40.

王仲. 燃气-蒸汽联合循环机组智能诊断与健康维护技术研究[D]. 北京: 华北电力大学, 2021.

Michael A N. Neural networks and deep learning[M]. [S.l.] :Determination Press, 2015:97.

BISHOP C M. Pattern recognition and machine learning[M]. [S.l.] :Springer,2006.

罗茂春. 基于FMI的航空发动机控制系统多学科联合仿真[D]. 南京: 南京航空航天大学, 2019.

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Forward Design of the Control System of Indigenously Developed Heavy-duty Gas Turbines

Study on Compound Impingement Cooling Characteristics of Ring Segment Components in a Heavy-duty Gas Turbine

Research on Error Correction Method for Two-dimensional Infrared Radiation Thermometry of Gas Turbine Combustor Wall

Key Technologies for Hydrogen-blended Natural Gas Power Generation Retrofit of Gas Turbine Power Plants

Numerical Simulation of Flow and Emission Characteristics in Hydrogen-blended Combustion of Gas Turbines

Related Author

ZHANG Xiaoyi

ZHANG Xuan

BAI Tao

Lü Xin

XIE Jinyu

SHEN Xinjun

YAN Binbin

YIN Yuyang

Related Institution

Department of Energy and Power Engineering, Tsinghua University

National Engineering Research Center of Power Generation Control and Safety, School of Energy and Environment, Southeast University

State Key Laboratory of Clean Energy Utilization, Zhejiang University

Shanghai Power Equipment Research Institute Co., Ltd.

School of Mechanical Engineering, Shanghai Jiao Tong University

AI问答

⁰