卷积神经网络的空调系统故障诊断可解释研究

doi:10.19784/j.cnki.issn1672-0172.2024.99.035

家电科技 ›› 2024, Vol. 0 ›› Issue (zk): 170-174.doi: 10.19784/j.cnki.issn1672-0172.2024.99.035

卷积神经网络的空调系统故障诊断可解释研究

熊成龙¹, 李冠男^1,2, 劳春峰³, 李伟³, 王东岳⁴, 代传民⁴, 李锟³

1.城市建设学院,武汉科技大学湖北武汉 430065;
2.湖北省城市更新工程研究中心,武汉科技大学湖北武汉 430065;
3.青岛海尔空调器有限总公司山东青岛 266101;
4.青岛海尔智能技术研发有限公司山东青岛 266101

出版日期:2024-12-10 发布日期:2024-12-31
通讯作者: 李冠男,E-mail：guannanli@wust.edu.cn。
作者简介:熊成龙,硕士学位。研究方向：空调系统的故障检测诊断研究。地址：武汉科技大学城市建设学院。E-mail：xcl0012@163.com。
基金资助:
国家自然科学基金项目（51906181）; 武汉科技大学“十四五”湖北省优势特色学科（群）项目（2023D0504,2023D0501）; 武汉科技大学研究生创新创业基金（JCX2023026）

Interpretation study on convolutional neural networks-based fault diagnosis of air conditioning system

XIONG Chenglong¹, LI Guannan^1,2, LAO Chunfeng³, LI Wei³, WANG Dongyue⁴, DAI Chuanmin⁴, LI Kun³

1. School of Urban Construction, Wuhan University of Science and Technology Wuhan 430065;
2. Hubei Provincial Engineering Research Center of Urban Regeneration, Wuhan University of Science and Technology Wuhan 430065;
3. Qingdao Haier Air Conditioner Gen.Corp., LTD. Qingdao 266101;
4. Qingdao Haier Smart Technology R&D CO.,LTD. Qingdao 266101

Online:2024-12-10 Published:2024-12-31

摘要/Abstract

摘要： 深度学习,特别是卷积神经网络（Convolutional neural networks,CNN）,在建筑能源系统受到广泛关注。在空气处理机组（Air handling unit,AHU）的故障诊断领域中,CNN的诊断性能的有效性和适用性需要进一步验证。同时,CNN故障诊断模型缺乏可解释性,不利于其在实际工程的推广应用。为解决以上问题,基于公开的ASHRAE RP-1312 AHU故障数据,利用CNN构建了故障诊断模型,同时基于逐层相关传播方法对CNN模型进行解释。结果表明,基于CNN的诊断模型展现出较好的适用性,平均诊断正确率为99.94%。逐层相关传播方法为CNN模型提供了良好的可解释性,识别出模型在决策过程中的诊断机制。最后,通过对卷积层数、学习率以及β参数方面深入分析模型参数对解释结果的影响。

关键词: 空气处理机组, 卷积神经网络, 逐层相关传播, 故障诊断, 可解释

Abstract: Deep learning, particularly convolutional neural networks (CNN), has garnered significant attention in the field of building energy systems. In the context of fault diagnosis for air handling units (AHU), the effectiveness and applicability of CNN's diagnostic performance require further validation. Additionally, the lack of interpretability in CNN fault diagnosis models hinders their broader application in practical engineering. To address these issues, utilized the publicly available ASHRAE RP-1312 AHU fault data to develop a fault diagnosis model based on CNN, and employed the layer-wise relevance propagation (LRP) method to interpret the CNN model. The results demonstrated that the CNN-based diagnostic model exhibits good applicability, with an average diagnostic accuracy of 99.94%. The LRP method provides strong interpretability for the CNN model, and identifies the diagnosis mechanism of the model in the decision-making process. Finally, an in-depth analysis was conducted on the impact of model parameters such as the number of convolutional layers, learning rate and β parameter on the interpretation results.

Key words: Air handling unit, Convolutional neural network, Fault diagnosis, Layer-wise relevance propagation, Interpretation

中图分类号:

TM925
TP277

熊成龙, 李冠男, 劳春峰, 李伟, 王东岳, 代传民, 李锟. 卷积神经网络的空调系统故障诊断可解释研究[J]. 家电科技, 2024, 0(zk): 170-174.

XIONG Chenglong, LI Guannan, LAO Chunfeng, LI Wei, WANG Dongyue, DAI Chuanmin, LI Kun. Interpretation study on convolutional neural networks-based fault diagnosis of air conditioning system[J]. Journal of Appliance Science & Technology, 2024, 0(zk): 170-174.

/ 推荐

参考文献 17

[1]	Ahmad M W, Mourshed M, Yuce B, et al.Computational intelligence techniques for HVAC systems: A review[J]. Building Simulation, 2016, 9: 359-398.
[2]	Yan K, Huang J, Shen W, et al.Unsupervised learning for fault detection and diagnosis of air handling units[J]. Energy & Buildings, 2020, 210: 109689.
[3]	Deshmukh S, Samouhos S, Glicksman L, et al.Fault detection in commercial building VAV AHU: A case study of an academic building[J]. Energy & Buildings, 2019, 201: 163-173.
[4]	Otter D W, Medina J R, Kalita J K.A Survey of the Usages of Deep Learning for Natural Language Processing[J]. IEEE transactions on neural networks and learning systems, 2021, 32: 604-624.
[5]	吴爽, 冯涛, 王晶. 基于迁移学习的室内小样本声源定位方法研究[J]. 家电科技, 2023(01): 74-78.
[6]	毛跃辉. 基于深度学习的语音识别技术研究及其在空调上的应用[J]. 家电科技, 2019(04): 54-58.
[7]	Li G, Yao Z, Chen L, et al.An interpretable graph convolutional neural network based fault diagnosis method for building energy systems[J]. Building Simulation, 2024, 17: 1113-1136.
[8]	Yan Y, Cai J, Tang Y, et al.Fault diagnosis of HVAC AHUs based on a BP-MTN classifier[J]. Building and Environment, 2023, 227: 109779.
[9]	Fan B, Du Z, Jin X, et al.A hybrid FDD strategy for local system of AHU based on artificial neural network and wavelet analysis (Article)[J]. Building and Environment, 2010, 45: 2698-2708.
[10]	Xiong C, Li G, Yan Y, et al.Effects of various information scenarios on layer-wise relevance propagation-based interpretable convolutional neural networks for air handling unit fault diagnosis[J]. Building Simulation, 2024: 1-22.
[11]	Rudin C, Chen C, Chen Z, et al.Interpretable Machine Learning: Fundamental Principles and 10 Grand Challenges[J]. Statistic Surveys, 2021.
[12]	Naser M Z.An engineer's guide to eXplainable Artificial Intelligence and Interpretable Machine Learning: Navigating causality, forced goodness, and the false perception of inference[J]. Automation in Construction, 2021, 129: 103821.
[13]	Preece A, Harborne D, Braines D, et al.Stakeholders in Explainable AI[J]. arXiv preprint arXiv, 2018.
[14]	Li S, Wen J.Development and Validation of a Dynamic Air Handling Unit Model, Part I[J]. ASHRAE Transactions, 2010, 116: 45-56.
[15]	Cheng H, Chen H, Li Z, et al.Ensemble 1-D CNN diagnosis model for VRF system refrigerant charge faults under heating condition[J]. Energy & Buildings, 2020, 224: 110256.
[16]	Zhao B, Zhang X, Zhan Z, et al.Deep multi-scale convolutional transfer learning network: A novel method for intelligent fault diagnosis of rolling bearings under variable working conditions and domains[J]. Neurocomputing, 2020, 407: 24-38.
[17]	Li H, Tian Y, Mueller K, et al.Beyond saliency: Understanding convolutional neural networks from saliency prediction on layer-wise relevance propagation[J]. Image and Vision Computing, 2019, 83: 70-86.

卷积神经网络的空调系统故障诊断可解释研究

Interpretation study on convolutional neural networks-based fault diagnosis of air conditioning system

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献 17

相关文章 3

Metrics

本文评价

推荐阅读 10

[1]	吴爽, 冯涛, 王晶. 基于迁移学习的室内小样本声源定位方法研究[J]. 家电科技, 2023, 0(1): 74-79.
[2]	刘圆圆, 江俊, 曹继来, 王利亚, 赵圣宇. 基于IVMD-SVM的冰箱声振故障诊断方法[J]. 家电科技, 2023, 0(1): 80-83.
[3]	赵洋, 张子祺, 汪超, 石里明, 杨荞慧. 小型制冷压缩机阀片防失效技术研究进展[J]. 家电科技, 2022, 0(4): 50-57.