Domain-Adversarial Transfer Learning-Driven Adaptive Diagnosis Method for Cross-Condition Multi-Faults in Subway Train Air Conditioning Systems

Zhang Tongle; Yang Chuang; Chen Huanxin; Zhang Xue

doi:10.12465/issn.0253-4339.20250614001

您当前的位置：

首页 >

文章列表页 >

Domain-Adversarial Transfer Learning-Driven Adaptive Diagnosis Method for Cross-Condition Multi-Faults in Subway Train Air Conditioning Systems

更新时间：2026-03-24

- Domain-Adversarial Transfer Learning-Driven Adaptive Diagnosis Method for Cross-Condition Multi-Faults in Subway Train Air Conditioning Systems
- Journal of Refrigeration Pages: 1-9(2026)
- 作者机构：
  
  1.华中科技大学中欧清洁与可再生能源学院武汉 430074
  2. 华中科技大学能源与动力工程学院武汉 430074
- 作者简介：
- 基金信息：
  
  the National Natural Science Foundation of China(51876070)
- DOI：10.12465/issn.0253-4339.20250614001
  CLC： U270.38⁺3;TP183;TP306⁺.3
- CSTR：XXXXX.XX.XXX.20250614001
- Received：14 June 2025，
  
  Revised：2025-09-04，
  
  Accepted：12 September 2025，
  
  Online First：24 March 2026，
- 稿件说明：
移动端阅览
张同乐,杨闯,陈焕新等.域对抗迁移学习驱动的地铁列车空调跨工况多故障自适应诊断方法[J].制冷学报,

Zhang Tongle,Yang Chuang,Chen Huanxin,et al.Domain-Adversarial Transfer Learning-Driven Adaptive Diagnosis Method for Cross-Condition Multi-Faults in Subway Train Air Conditioning Systems[J].Journal of Refrigeration,
张同乐,杨闯,陈焕新等.域对抗迁移学习驱动的地铁列车空调跨工况多故障自适应诊断方法[J].制冷学报, DOI：10.12465/issn.0253-4339.20250614001. CSTR： XXXXX.XX.XXX.20250614001.

Zhang Tongle,Yang Chuang,Chen Huanxin,et al.Domain-Adversarial Transfer Learning-Driven Adaptive Diagnosis Method for Cross-Condition Multi-Faults in Subway Train Air Conditioning Systems[J].Journal of Refrigeration, DOI：10.12465/issn.0253-4339.20250614001. CSTR： XXXXX.XX.XXX.20250614001.

摘要

地铁空调系统的高效故障诊断对降低能耗和保障乘客舒适度至关重要。本文针对地铁空调运行中特征分布差异大、故障复杂的问题，提出一种基于域对抗神经网络（DANN）的无监督迁移学习故障诊断方法。在多功能综合试验车中采集了冷凝器结垢、通风系统结垢和制冷剂泄漏3种单故障及3种单故障组合的并发故障数据，在不同工况和压缩机频率下进行迁移学习验证，单故障诊断准确率达97.30%~98.90%，并发故障诊断准确率为77.80%~86.70%。使用UMAP非线性降维法与SHAP模型可解释性工具探究了并发故障诊断精度低于单故障诊断精度的深层原理。将DANN与其他2种迁移学习模型对比，波动幅度显著小于对比模型。在特征分布差异较大的迁移学习诊断任务中，诊断准确率波动幅度仅1%，在并发故障特征重叠时仍保持较高诊断精度。

Abstract

Efficient fault diagnosis of metro air-conditioning systems is essential for reducing energy consumption and ensuring passenger comfort. This paper proposes an unsupervised transfer learning method based on a domain-adversarial neural network （DANN） to address the challenges involving diverse feature distributions and complex faults. Data obtained from three single faults （condenser fouling， ventilation fouling， and refrigerant leakage） and their concurrent combinations are collected from a multifunctional test vehicle under various operating conditions and compressor frequencies. The DANN achieved accuracy values ranging from 97.30%-98.90% for single faults and 77.80%-86.70% for concurrent faults. Uniform Manifold Approximation and Projection （UMAP） and SHapley Additive exPlanations （SHAP） analyses revealed the underlying reasons for the less accurate concurrent fault diagnosis. Compared with the two baseline transfer learning models， DANN exhibited markedly smaller performance fluctuations， maintaining high accuracy even under large feature distribution shifts and overlapping concurrent fault features.

关键词

Keywords

references

Fan Xianwang ， Zhang Huan ， Liu Jiali ， et al . Numerical analysis of the subway tunnel thermal environment to predict the train-mounted condenser inlet temperature in the cold climate zone of China ［J］. Energy and Buildings ， 2024 ， 324 ： 114928 .

中国城市轨道交通协会 . 城市轨道交通2023年度统计和分析报告［EB/OL］. （ 2024-03-29 ）［ 2025-06-05 ］. https：//www.camet.org.cn/xytj/tjxx/14894.shtml https://www.camet.org.cn/xytj/tjxx/14894.shtml .

China Urban Rail Transit Association . Urban Rail Transit 2023 Annual Statistics and Analysis Report ［EB/OL］.（ 2024-03-29 ）［ 2025-06-05 ］. https：//www.camet.org.cn/xytj/tjxx/14894.shtml https://www.camet.org.cn/xytj/tjxx/14894.shtml .

Guan Bowen ， Liu Xiaohua ， Zhang Tao ， et al . Energy consumption of subway stations in China： Data and influencing factors ［J］. Sustainable Cities and Society ， 2018 ， 43 ： 451 - 461 .

Han H ， Gu B ， Kang J ， et al . Study on a hybrid SVM model for chiller FDD applications ［J］. Applied Thermal Engineering ， 2011 ， 31 （ 4 ）： 582 - 592 .

周璇，王晓佩，梁列全，等 . 基于随机森林算法的制冷剂充注量故障诊断［J］. 华南理工大学学报（自然科学版）， 2020 ， 48 （ 2 ）： 16 - 24 .

Zhou Xuan ， Wang Xiaopei ， Liang Liequan ， et al . Random forests algorithm-based fault diagnosis for refrigerant charge ［J］. Journal of South China University of Technology （Natural Science Edition）， 2020 ， 48 （ 2 ）： 16 - 24 .

Sun Shaobo ， Li Guannan ， Chen Huanxin ， et al . A hybrid ICA-BPNN-based FDD strategy for refrigerant charge faults in variable refrigerant flow system ［J］. Applied Thermal Engineering ， 2017 ， 127 ： 718 - 728 .

刘飞天，韩华，杨钰婷，等 . 基于CNN与ResNet的离心式冷水机组故障诊断［J］. 暖通空调， 2023 ， 53 （ 9 ）： 50 - 56 .

Liu Feitian ， Han Hua ， Yang Yuting ， et al . Fault diagnosis of centrifugal chillers based on CNN and ResNet ［J］. Journal of HV&AC ， 2023 ， 53 （ 9 ）： 50 - 56 .

Yan Ke ， Chen Xinke ， Zhou Xiaokang ， et al . Physical model informed fault detection and diagnosis of air handling units based on transformer generative adversarial network ［J］. IEEE Transactions on Industrial Informatics ， 2023 ， 19 （ 2 ）： 2192 - 2199 .

Jiang Minhui ， Chen Huanxin ， Yang Chuang . A metro train air conditioning system fault diagnosis method based on explainable artificial intelligence： Considering interpretability and generalization ［J］. International Journal of Refrigeration ， 2025 ， 174 ： 47 - 59 .

Zhou Yuanlong ， Bi Haiquan ， Wang Honglin ， et al . Time-delay characteristics of air-conditioning system for subway trains ［J］. Journal of Building Engineering ， 2021 ， 40 ： 102731 .

吴楠楠，赵亮，蒋志鹏，等 . 不同气候区地铁列车空调全年动态负荷特性［J］. 制冷， 2023 ， 42 （ 1 ）： 43 - 48 .

Wu Nannan ， Zhao Liang ， Jiang Zhipeng ， et al . Analyze on annual load distribution characteristics of metro vehicles air conditioning in different climatic zones ［J］. Refrigeration ， 2023 ， 42 （ 1 ）： 43 - 48 .

刘俊媛 . 地铁区间隧道传热特性研究［D］. 武汉：华中科技大学， 2022 .

Liu Junyuan . Heat transfer characteristics of subway interval tunnels ［D］. Wuhan ： Huazhong University of Science and Technology ， 2022 .

Ganin Y ， Lempitsky V . Unsupervised domain adaptation by backpropagation ［C］// The 32nd International Conference on Machine Learning ， 2015 ： 1180 - 1189 .

刘玉琳，白杨，崔斌，等 . 面向机器学习的自动化特征工程研究综述［J］. 计算机应用与软件， 2025 ， 42 （ 1 ）： 1 - 10 .

Liu Yulin ， Bai Yang ， Cui Bin ， et al . A survey on automated feature engineering for machine learning ［J］. Computer Applications and Software ， 2025 ， 42 （ 1 ）： 1 - 10 .

Ben-David S ， Blitzer J ， Crammer K ， et al . Analysis of representations for domain adaptation ［M］// Advances in Neural Information Processing Systems 19 . Cambridge， MA， USA ： The MIT Press ， 2007 ： 137 - 144 .

李昕璐，欧阳瑞志，高帝逸，等 . 基于时序UMAP和强化学习优化随机森林的碳价格预测模型［J］. 中国高新科技， 2025 （ 14 ）： 137 - 139 .

Li Xinlu ， Ouyang Ruizhi ， Gao Diyi ， et al . Carbon price forecasting using temporal UMAP and reinforcement learning-optimized random forest ［J］. China High-Tech ， 2025 （ 14 ）： 137 - 139 .

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Evidence-Driven and Neural Network-Based Fault Monitoring and Diagnosis Technology for District Cooling and Heating System

Interpretability Study on the Fault-Diagnosis Model of the Heat Pipe / Vapor-Compression Composite Air-Conditioning System

Research Status and Prospects of Terminals of Central Air-Conditioning Systems

Related Author

Xu Baoping

Hou Hongjuan

Wang Xi

Cen Xiaotong

Zhang Yiqi

Huang Shuoquan

Li Xiuming

Di Yanqiang

Related Institution

State Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources， North China Electric Power University

School of Energy Power and Mechanical Engineering， North China Electric Power University

School of Mechanical Engineering， Beijing Institute of Technology

China Academy of Building Research

Energy Saving and Low-Carbon Technology of Process Industry Engineering Research Center of Liaoning Province， Northeastern University

Address：10/F, Yindu Mansion, 67 Fucheng Road, Haidian Dist, Beijing, China Postal code：100142
Tel：010-68711412 Email：editor@car.org.cn
Technical support is provided by Beijing Founder electronics co., LTD 京ICP备09064830号-19 京公网安备11010802024621
It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.

⁰