基于CNN-LSTM的换热器污垢因子预测研究

doi:10.11949/0438-1157.20241137

摘要/Abstract

摘要：

对换热器的结垢状态进行精准预测，可以及时了解结垢程度，从而有针对性地实施清洁，对提高换热器使用经济性和生产安全性具有重要意义。利用卷积神经网络（CNN）与长短期记忆网络（LSTM）的模型集成对换热器污垢因子进行预测，采用大量换热器历史数据对所建立的CNN-LSTM模型进行训练，取得了良好的预测效果。与单一的CNN和LSTM模型及文献中的多层感知器神经网络模型（MLPNN）相比，所建立的CNN-LSTM模型具有准确性更高、稳定性更强的特点。在所展示的案例中，决定系数（R²）为0.98167，平均绝对百分误差（MAPE）为3.199×10^-3，不仅为解决换热器结垢问题提供了理论基础，而且为换热器安全运行与维护策略提供了更科学、精准的依据，有助于提升整个换热工段的经济性和生产安全性。

关键词: 换热器, 模型, 预测, 神经网络, 集成, 算法

Abstract:

Accurate prediction of the fouling state of the heat exchanger can timely understand the degree of fouling, so as to implement targeted cleaning, which is of great significance to improving the economic use and production safety of the heat exchanger. The model integration of convolutional neural network (CNN) and long short-term memory network (LSTM) was used to predict the heat exchanger fouling factor. A large number of historical data of heat exchanger are used to train the CNN-LSTM model, and excellent prediction results are obtained. Compared with the single CNN and LSTM models and the multi-layer perceptron neural network (MLPNN) models in the literature, the CNN-LSTM model is more accurate and more stable. In the case shown, the coefficient of determination (R²) is 0.98167, and the average absolute percentage error (MAPE) is 3.199×10^-3. The establishment of the model not only provides a theoretical basis for solving the scale problem of the heat exchanger, but also provides a more scientific and accurate basis for the safe operation and maintenance strategy of the heat exchanger. It helps to improve the economy and production safety of the entire heat exchange section.

Key words: heat exchanger, model, prediction, neural networks, integration, algorithm

中图分类号:

TK 172

张晗筱, 王瑞琪, 张亚婷. 基于CNN-LSTM的换热器污垢因子预测研究[J]. 化工学报, 2025, 76(4): 1671-1679.

Hanxiao ZHANG, Ruiqi WANG, Yating ZHANG. Prediction of scale factor of heat exchangers based on CNN-LSTM neural network[J]. CIESC Journal, 2025, 76(4): 1671-1679.

图/表 11

图1 技术路线

Fig.1 Technology roadmap

图2 相关性矩阵热力图

Fig.2 Correlation matrix heat map

图3 数据清洗对比

Fig.3 Comparison of data cleaning

表1 归一化处理对比

Table 1 Normalized processing comparison

归一化方式	R²
[-2,2]归一化	0.90955
[-2,0]归一化	0.65239
[-1,0]归一化	0.73828
[-1,1]归一化	0.91915
[0,1]归一化	0.95739
[0,2]归一化	0.88683
不进行归一化	0.61451

图4 所提出的CNN-LSTM模型的架构

Fig.4 Architecture of the CNN-LSTM model proposed

图5 便携式污垢研究装置

Fig.5 Portable fouling research unit

图6 CNN-LSTM模型初始预测结果

Fig.6 CNN-LSTM model initial prediction results

表2 所得到的神经网络在训练集和测试集的性能

Table 2 The performance of the obtained neural network on the training set and the testing set

评价指标	训练集	测试集	所有数据集
R²	0.98441	0.98228	0.98167
MAE	1.1246×10^-2	3.7923×10^-2	—
MAPE	0.29784×10^-3	3.1992×10^-3	—

图7 CNN-LSTM预测结果

Fig.7 CNN-LSTM model prediction results

表3 单变量预测结果对比

Table 3 Comparison of univariate prediction results

单变量	训练集R²	测试集R²	所有数据集R²
密度	0.92198	0.91917	0.92295
流体氧含量	0.92218	0.91054	0.91925
流体速度	0.9222	0.73436	0.84173
流体温度	0.9226	0.91167	0.91998
等效直径	0.92204	0.78172	0.86249
表面温度	0.93016	0.93329	0.93351
时间	0.93453	0.88835	0.91606

表4 CNN-LSTM模型与其他模型预测对比

Table 4 Prediction comparison between CNN-LSTM model and other models

模型	R²
MLPNN	0.97782
CNN	0.82476
LSTM	0.65960
CNN-LSTM	0.98167

参考文献 37

1	Muthukrishnan S, Krishnaswamy H, Thanikodi S, et al. Support vector machine for modelling and simulation of heat exchangers[J]. Thermal Science, 2020, 24(1 Part B): 499-503.
2	Sadeghianjahromi A, Wang C C. Heat transfer enhancement in fin-and-tube heat exchangers—a review on different mechanisms[J]. Renewable and Sustainable Energy Reviews, 2021, 137: 110470.
3	赵洪德, 刘继东, 周广启, 等. 换热器在石油化工行业中的应用及维护[J]. 造纸装备及材料, 2022, 51(3): 52-54.
	Zhao H D, Liu J D, Zhou G Q, et al. Application and maintenance of heat exchanger in petrochemical industry[J]. Papermaking Equipment & Materials, 2022, 51(3): 52-54.
4	田佳阳, 贾林权, 王彧斐, 等. 考虑关键换热器备用的原油预热系统清垢周期优化[J]. 化工学报, 2016, 67(12): 5183-5189.
	Tian J Y, Jia L Q, Wang Y F, et al. Schedule optimization for fouling removal in refinery heat exchanger networks with backup unit[J]. CIESC Journal, 2016, 67(12): 5183-5189.
5	Tian J Y, Wang Y F, Feng X. Simultaneous optimization of flow velocity and cleaning schedule for mitigating fouling in refinery heat exchanger networks[J]. Energy, 2016, 109: 1118-1129.
6	Davoudi E, Vaferi B. Applying artificial neural networks for systematic estimation of degree of fouling in heat exchangers[J]. Chemical Engineering Research and Design, 2018, 130: 138-153.
7	刘子奇, 赵轩, 王洋, 等. 热泵站换热器污垢研究[J]. 无线互联科技, 2018, 15(17): 69-70.
	Liu Z Q, Zhao X, Wang Y, et al. Research on heat exchanger fouling of heat pump station[J]. Wireless Internet Technology, 2018, 15(17): 69-70.
8	Polley G T, Wilson D I, Yeap B L, et al. Evaluation of laboratory crude oil threshold fouling data for application to refinery pre-heat trains[J]. Applied Thermal Engineering, 2002, 22(7): 777-788.
9	Yeap B L, Wilson D I, Polley G T, et al. Mitigation of crude oil refinery heat exchanger fouling through retrofits based on thermo-hydraulic fouling models[J]. Chemical Engineering Research and Design, 2004, 82(1): 53-71.
10	Łopata S, Ocłoń P. Numerical study of the effect of fouling on local heat transfer conditions in a high-temperature fin-and-tube heat exchanger[J]. Energy, 2015, 92: 100-116.
11	徐源, 陶苗苗, 孙灵芳. 基于灰色神经网络的换热器污垢预测研究[J]. 吉林农业科技学院学报, 2014, 23(1): 45-48.
	Xu Y, Tao M M, Sun L F. Research on fouling prediction of heat exchanger based on grey theory and neural network[J]. Journal of Jilin Agricultural Science and Technology University, 2014, 23(1): 45-48.
12	余文敏, 余刃, 毛伟, 等. 基于SWLSTM的换热器在线性能预测[J]. 舰船科学技术, 2023, 45(21): 153-157.
	Yu W M, Yu R, Mao W, et al. On-line performance prediction of heat exchanger based on SWLSTM[J]. Ship Science and Technology, 2023, 45(21): 153-157.
13	Mohanty D K, Singru P M. Fouling analysis of a shell and tube heat exchanger using local linear wavelet neural network[J]. International Journal of Heat and Mass Transfer, 2014, 77: 946-955.
14	刘新, 伍俊楠, 潘殿琦, 等. 基于改进FMEA的污水厂设备故障风险评估模型[J]. 中国安全科学学报, 2024, 34(8): 101-107.
	Liu X, Wu J N, Pan D Q, et al. Equipment failure risk assessment model of wastewater treatment plant based on improved FMEA[J]. China Safety Science Journal, 2024, 34(8): 101-107.
15	周怡, 彭国文, 黄召, 等. 基于SBAS-InSAR和BPNN的铀尾矿坝形变智能监测与预测[J]. 中国安全科学学报, 2024, 34(4): 145-152.
	Zhou Y, Peng G W, Huang Z, et al. Intelligent monitoring and prediction of deformation of uranium tailings dam based on SBAS-InSAR and BPNN[J]. China Safety Science Journal, 2024, 34(4): 145-152.
16	翁润滢, 孙斌, 赵玉晓, 等. 基于自适应最优核和卷积神经网络的气液两相流流型识别方法[J]. 化工学报, 2018, 69(12): 5065-5072.
	Weng R Y, Sun B, Zhao Y X, et al. Flow pattern recognition method of gas-liquid two-phase flow based on adaptive optimal kernel and convolution neural network[J]. CIESC Journal, 2018, 69(12): 5065-5072.
17	Wu H, Zhao J S. Deep convolutional neural network model based chemical process fault diagnosis[J]. Computers & Chemical Engineering, 2018, 115: 185-197.
18	孙先亮, 李健, 韩哲哲, 等. 基于数据驱动的卷积神经网络电容层析成像图像重建[J]. 化工学报, 2020, 71(5): 2004-2016.
	Sun X L, Li J, Han Z Z, et al. Data-driven image reconstruction of electrical capacitance tomography based on convolutional neural network[J]. CIESC Journal, 2020, 71(5): 2004-2016.
19	王路瑶, 吴斌, 杜志敏, 等. 基于长短期记忆神经网络的数据中心空调系统传感器故障诊断[J]. 化工学报, 2018, 69(S2): 252-259.
	Wang L Y, Wu B, Du Z M, et al. Sensor fault detection and diagnosis for data center air conditioning system based on LSTM neural network[J]. CIESC Journal, 2018, 69(S2): 252-259.
20	Rao R V, Saroj A. Economic optimization of shell-and-tube heat exchanger using Jaya algorithm with maintenance consideration[J]. Applied Thermal Engineering, 2017, 116: 473-487.
21	Asomaning S. The role of olefins in fouling of heat exchangers[D]. Vancouver: University of British Columbia, 1990.
22	Asomaning S. Heat exchanger fouling by petroleum asphaltenes[D]. Vancouver: University of British Columbia, 1997.
23	Sundaram B N. The effects of oxygen on synthetic crude oil fouling[D]. Vancouver: University of British Columbia, 1998.
24	Watkinson A P. Particulate fouling of sensible heat exchangers[D]. Vancouver: University of British Columbia, 1968.
25	Srinivasan M. Heat exchanger fouling of some Canadian crude oils[D]. Vancouver: University of British Columbia, 2008.
26	Feiz G. Annular heating probes in oil fouling: effects of wall shear stress[D]. Vancouver: University of British Columbia, 2015.
27	李义, 刘颖慰, 都健, 等. 基于大数据建模的玉米淀粉糖工艺参数预测系统开发[J]. 当代化工研究, 2022(14): 145-152.
	Li Y, Liu Y W, Du J, et al. System development for the prediction of key parameters in corn starch sugar process based on big data modeling[J]. Modern Chemical Research, 2022(14): 145-152.
28	郜亚东, 李冠, 赵立峰, 等. 一种四分位数法优化的GM(1,1)模型在高危边坡监测中应用研究[J]. 城市勘测, 2023(3): 171-175.
	Gao Y D, Li G, Zhao L F, et al. Application of GM (1,1) model optimized by quartile method in high risk slope monitoring[J]. Urban Geotechnical Investigation & Surveying, 2023(3): 171-175.
29	杜林颖, 于鸿彬, 侯立国, 等. 改进的BP神经网络对飞机换热器结垢厚度预测[J]. 计算机仿真, 2020, 37(1): 27-30.
	Du L Y, Yu H B, Hou L G, et al. Prediction of fouling thickness of aircraft heat exchanger by modified BP neural network[J]. Computer Simulation, 2020, 37(1): 27-30.
30	Zha W S, Liu Y P, Wan Y J, et al. Forecasting monthly gas field production based on the CNN-LSTM model[J]. Energy, 2022, 260: 124889.
31	Garcia C I, Grasso F, Luchetta A, et al. A comparison of power quality disturbance detection and classification methods using CNN, LSTM and CNN-LSTM[J]. Applied Sciences, 2020, 10(19): 6755.
32	张术琳, 张亚楠, 田超, 等. 基于CNN的化工园区火灾火焰图像识别研究[J]. 中国安全科学学报, 2024, 34(1): 179-186.
	Zhang S L, Zhang Y N, Tian C, et al. Study on flame image recognition of chemical industrial park fires based on convolutional neural network[J]. China Safety Science Journal, 2024, 34(1): 179-186.
33	Wang J, Sun L, Li H, et al. Prediction model of fouling thickness of heat exchanger based on TA-LSTM structure[J]. Processes, 2023, 11(9): 2594.
34	Barzegar R, Aalami M T, Adamowski J. Short-term water quality variable prediction using a hybrid CNN-LSTM deep learning model[J]. Stochastic Environmental Research and Risk Assessment, 2020, 34(2): 415-433.
35	吴婷婷, 许晓东, 吴云龙. 卷积神经网络中SPReLU激活函数的优化研究[J]. 计算机与数字工程, 2021, 49(8): 1637-1641.
	Wu T T, Xu X D, Wu Y L. Research on optimization of SPReLU activation function in convolutional neural network[J]. Computer & Digital Engineering, 2021, 49(8): 1637-1641.
36	Kim T Y, Cho S B. Predicting residential energy consumption using CNN-LSTM neural networks[J]. Energy, 2019, 182: 72-81.
37	Hosseini S, Khandakar A, Chowdhury M E H, et al. Novel and robust machine learning approach for estimating the fouling factor in heat exchangers[J]. Energy Reports, 2022, 8: 8767-8776.

[1]	产文, 余万, 王岗, 苏华山, 黄芬霞, 胡涛. 改进回热布局的Allam循环热力、经济性能分析和双目标优化[J]. 化工学报, 2025, 76(4): 1680-1692.
[2]	徐芳萍, 杨辉, 陈俊, 朱建勇, 陆荣秀. 基于迁移学习与残差注意力卷积网络的稀土元素组分含量软测量[J]. 化工学报, 2025, 76(4): 1647-1660.
[3]	王法正, 隋璘, 熊伟丽. 面向多采样率数据的TTPA-LSTM软测量建模[J]. 化工学报, 2025, 76(4): 1635-1646.
[4]	周洋, 梁聃, 汪恺, 张艺蓝, 贾立. 基于Alopex进化算法和集成学习的多模态神经网络的污水反渗透过程建模[J]. 化工学报, 2025, 76(4): 1661-1670.
[5]	田浩辰, 马志先, 王之浩. R1234ze（E）水平三维肋管外膜状凝结特性实验研究[J]. 化工学报, 2025, 76(3): 975-984.
[6]	侯亚祺, 张玮, 张鸿, 高飞雨, 胡嘉华. 基于机器学习与粒子群算法的LBM多相流模型优化[J]. 化工学报, 2025, 76(3): 1120-1132.
[7]	赵丽文, 刘桂莲. 基于系统集成的复杂催化反应系统性能强化及参数优化[J]. 化工学报, 2025, 76(3): 1111-1119.
[8]	郑欣雨, 任泽华, 周利, 柴士阳, 吉旭. 基于晶体图卷积神经网络的晶格能回归模型[J]. 化工学报, 2025, 76(3): 1084-1092.
[9]	姚国家, 王志, 苏昂, 冯东阁, 唐宏, 孙灵芳. 空气系数对煤粉预热解燃烧特性的影响分析[J]. 化工学报, 2025, 76(3): 1243-1252.
[10]	王大芬, 唐莉丽, 张鑫焱, 聂春雨, 李明珠, 吴菁. 基于时差的多输出tri-training异构软测量建模[J]. 化工学报, 2025, 76(3): 1143-1155.
[11]	王宗廷, 王丽丽, 孙晓岩, 夏力, 陶少辉, 项曙光. 基于简化相平衡关联式的高效简捷精馏塔模型[J]. 化工学报, 2025, 76(3): 1133-1142.
[12]	李京润, 杨思宇, 刘庆辉, 潘安, 王嘉岳, 符小贵, 余皓. 大规模风电耦合火电制氢多情景下不同运行策略分析[J]. 化工学报, 2025, 76(3): 1191-1206.
[13]	张恒, 魁殿禄, 常虹, 詹志刚. 机械应力对气体扩散层界面传输特性影响[J]. 化工学报, 2025, 76(2): 637-644.
[14]	尤潇楠, 范小强, 杨遥, 王靖岱, 阳永荣. 超临界乙烯和高压聚乙烯混合物的减压分离过程建模方法[J]. 化工学报, 2025, 76(2): 695-706.
[15]	党法璐, 孙志国, 高照, 王刚, 陈政宇, 张霖宙, 连竞存, 刘美佳, 张忠东, 刘超伟. 原油一步法催化裂解制低碳烯烃：实验和反应路径研究[J]. 化工学报, 2025, 76(2): 667-685.