Physical Anchor Residual-Driven Intelligent Soft Sensing for Amine Defoaming

Li Jiaqi; Lan Zhong; Hao Xiaojing; Cao Ying; Zhang Yunlu

doi:10.16085/j.issn.1000-6613.2026-0989

您当前的位置：

首页 >

文章列表页 >

Physical Anchor Residual-Driven Intelligent Soft Sensing for Amine Defoaming

更新时间：2026-08-20

- Physical Anchor Residual-Driven Intelligent Soft Sensing for Amine Defoaming
- Chemical Industry and Engineering Progress (2026)
- 作者机构：
  
  大连理工大学化工学院，辽宁大连 116031
- 作者简介：
- 基金信息：
- DOI：10.16085/j.issn.1000-6613.2026-0989
  CLC： TQ 02;TP183
- Received：19 June 2026，
  
  Revised：2026-08-10，
  
  Accepted：18 August 2026，
- 稿件说明：
移动端阅览
Li Jiaqi, Lan Zhong, Hao Xiaojing, et al. Physical Anchor Residual-Driven Intelligent Soft Sensing for Amine Defoaming[J/OL]. Chemical Industry and Engineering Progress, 2026.
DOI：

Li Jiaqi, Lan Zhong, Hao Xiaojing, et al. Physical Anchor Residual-Driven Intelligent Soft Sensing for Amine Defoaming[J/OL]. Chemical Industry and Engineering Progress, 2026. DOI： 10.16085/j.issn.1000-6613.2026-0989.

摘要

胺液吸收塔在污油回炼过程中频繁发泡，严重影响装置平稳运行。本文提出一种融合时间卷积网络（TCN）、双向长短期记忆网络（BiLSTM）、自适应注意力及物理锚点残差的复合软测量模型。该模型将预测目标从泡沫的绝对高度改为残差增量，以此缓解强自回归场景下的过度平滑和时滞问题。模型前端用TCN做多尺度空间解耦，用BiLSTM捕捉消泡剂降解的长周期依赖，注意力层聚焦突变时刻，最后通过双向极值钳制损失函数强化极值区间的预测精度。在发泡动力学仿真数据上的测试结果为：决定系数R²达0.9456，均方误差MSE相比LSTM基线降低19.5%。消融实验分析，物理锚点残差范式向模型注入强归纳偏置，为胺液吸收塔发泡过程的精准预测与安全调控提供了可解释的智能策略。

Abstract

Frequent foaming in amine absorption towers during waste-oil recycling seriously affects the stable and safe operation of refining units. To achieve accurate soft sensing of foam height under strong inertia

nonlinear disturbances

and defoamer degradation

this study proposed a hybrid model driven by physical-anchor residual learning. Instead of directly predicting the absolute foam height

the previous foam height was introduced as a physical anchor

and the network was trained to learn the residual increment caused by gas-flow fluctuation

impurity accumulation

and defoamer attenuation. This design reduced the dominance of strong autoregressive trends and alleviated over-smoothing and time-delay problems in dynamic prediction. The proposed framework integrated a Temporal Convolutional Network (TCN)

a Bidirectional Long Short-Term Memory (BiLSTM) network

an adaptive attention mechanism

and a Dual-Extreme Loss function. The TCN module extracted and decoupled multi-scale features from noisy process variables

while BiLSTM captured long-term dynamic dependencies related to defoamer degradation. The attention mechanism enhanced the response to abrupt operating changes

and the Dual-Extreme Loss strengthened prediction accuracy in both high-risk high-foam regions and low-foam defoaming regions. Tests on foaming-kinetics simulation data showed that the proposed residual TCN-BiLSTM-Attention model achieved a coefficient of determination (R²) of 0.9456 and a mean squared error (MSE) of 2.760

reducing MSE by 19.5% compared with the Long Short-Term Memory (LSTM) baseline. Ablation experiments demonstrated that the physical-anchor residual paradigm introduced a strong inductive bias

suppressed overfitting in limited-sample conditions

and enabled the deep network to focus on disturbance-induced increments. The proposed method provides an interpretable and reliable intelligent strategy for foam prediction

early warning

and safety-oriented control in amine absorption towers.

关键词

Keywords

references

王龙 , 廖洪波 , 李玉冰 , 等 . 废油资源化标准体系现状、存在问题和建设建议 [J ] . 中国标准化 , 2023 ( 14 ): 106 - 109 .

Wang L , Liao H B , Li Y B , et al . Current situation, existing problems, and construction suggestions of waste oil recovery standards system [J ] . China Standardization , 2023 ( 14 ): 106 - 109 .

余燕燕 , 杨欣 . 废油再生现状及其工艺技术简介 [J ] . 资源再生 , 2019 ( 6 ): 46 - 49 .

Yu Y Y , Yang X . Introduction to waste oil regeneration status and process technology [J ] . Resource Recycling , 2019 ( 6 ): 46 - 49 .

柳艳青 , 杨金生 , 王象会 , 等 . 炼厂脱硫废剂处理试验研究 [J ] . 石化技术 , 2022 , 29 ( 5 ): 105 - 106 .

Liu Y Q , Yang J S , Wang X H , et al . Brief analysis on the experimental results of treating waste desulfurizing agent in Refinery [J ] . Petrochemical Industry Technology , 2022 , 29 ( 5 ): 105 - 106 .

刘杰 . 废矿物油回收再生利用技术及其实践方案分析 [J ] . 现代盐化工 , 2022 , 49 ( 6 ): 43 - 45 .

Liu J . Analysis of recycling technology of waste mineral oil and its practical scheme [J ] . Jiangsu Salt Science & Technology , 2022 , 49 ( 6 ): 43 - 45 .

李明一 . 炼油企业胺液脱硫系统优化与节能技术 [J ] . 石油石化绿色低碳 , 2025 , 10 ( 5 ): 22 - 26 .

Li M Y . Optimization and energy-saving technology of amine liquid desulfurization system in petroleum refinery [J ] . Energy Conservation and Emission Reduction in Petroleum and Petrochemical Industry , 2025 , 10 ( 5 ): 22 - 26 .

刘健 , 王鹏 , 丁棚格 , 等 . 炼油厂胺液再生装置节能优化措施应用 [J ] . 石油石化绿色低碳 , 2024 , 9 ( 5 ): 44 - 48 .

Liu J , Wang P , Ding P G , et al . Application of energy-saving optimization measures for MDEA solvent regeneration unit in refineries [J ] . Energy Conservation and Emission Reduction in Petroleum and Petrochemical Industry , 2024 , 9 ( 5 ): 44 - 48 .

蒋浩 , 陈丙珍 . 化工过程稳定性分析研究进展 [J ] . 化工学报 , 2018 , 69 ( 1 ): 76 - 87 .

Jiang H , Chen B Z . Research progress of chemical process stability analysis [J ] . CIESC Journal , 2018 , 69 ( 1 ): 76 - 87 .

张宁峰 . MDEA脱硫胺液发泡原因分析及对策 [J ] . 石油化工应用 , 2020 , 39 ( 3 ): 116 - 118 .

Zhang N F . Cause analysis and countermeasures of foaming in MDEA desulfurization amine solution [J ] . Petrochemical Industry Application , 2020 , 39 ( 3 ): 116 - 118 .

高旭晖 , 崔明皓 , 郝天臻 . 导致胺液严重发泡的主要原因研究 [J ] . 化学工程与装备 , 2025 ,( 12 ): 23 - 26 .

赵长斌 , 张连怀 . 胺液脱硫系统顽固性发泡的原因及对策 [J ] . 石化技术与应用 , 2024 , 42 ( 1 ): 60 - 63 .

Zhao C B , Zhang L H . Causes and countermeasures of refractory foaming in amine solution desulphurization system [J ] . Petrochemical Technology & Application , 2024 , 42 ( 1 ): 60 - 63 .

鲁亚青 , 张楠 , 王文宗 , 等 . 聚醚改性硅油消泡剂消泡机理研究 [J ] . 有机硅材料 , 2010 , 24 ( 2 ): 80 - 84 .

Lu Y Q , Zhang N , Wang W Z , et al . Study on defoaming mechanism of polyether modified polysiloxane compounds [J ] . Silicone Material , 2010 , 24 ( 2 ): 80 - 84 .

陈嵩嵩 , 鲍艾丽 , 霍锋 , 等 . 人工智能(AI)在复杂化工过程设计中的应用: 现状、挑战与展望 [J ] . 化工进展 , 2025 , 44 ( 8 ): 4821 - 4837 .

Chen S S , Bao A L , Huo F , et al . Application of artificial intelligence(AI)in the design of complex chemical engineering processes: Status, challenges and prospects [J ] . Chemical Industry and Engineering Progress , 2025 , 44 ( 8 ): 4821 - 4837 .

胡云洋 . 人工智能在化工生产过程自动化中的应用 [J ] . 塑料工业 , 2025 , 53 ( 3 ): 192 .

Hu Y Y . Application of artificial intelligence in automation of chemical production process [J ] . China Plastics Industry , 2025 , 53 ( 3 ): 192 .

Wang X , Qi X M , Zhang Y . Deep learning with local spatiotemporal structure preserving for soft sensor development of complex industrial processes [J ] . Applied Soft Computing , 2024 , 164 : 111974 .

Sun Q Q , Ge Z Q . A survey on deep learning for data-driven soft sensors [J ] . IEEE Transactions on Industrial Informatics , 2021 , 17 ( 9 ): 5853 - 5866 .

李文华 , 叶洪涛 , 罗文广 , 等 . 基于MHSA-LSTM的软测量建模及其在化工过程中的应用 [J ] . 化工学报 , 2024 , 75 ( 12 ): 4654 - 4665 .

Li W H , Ye H T , Luo W G , et al . Soft sensor modeling based on MHSA-LSTM and its application in chemical process [J ] . CIESC Journal , 2024 , 75 ( 12 ): 4654 - 4665 .

Perera Y S , Ratnaweera D A A C , Dasanayaka C H , et al . The role of artificial intelligence-driven soft sensors in advanced sustainable process industries: A critical review [J ] . Engineering Applications of Artificial Intelligence , 2023 , 121 : 105988 .

Zhang L , Ren G F , Li S L , et al . A novel soft sensor approach for industrial quality prediction based TCN with spatial and temporal attention [J ] . Chemometrics and Intelligent Laboratory Systems , 2025 , 257 : 105272 .

李越 . 面向化工动态过程预测的深度学习建模研究 [D ] . 重庆 : 重庆大学 , 2024 .

Li Y . Research on deep learning modeling for chemical dynamic process prediction [D ] . Chongqing : Chongqing University , 2024 .

Xie C R , Yao R J , Zhu L Y , et al . Soft-sensor development through deep learning with spatial and temporal feature extraction of complex processes [J ] . Industrial & Engineering Chemistry Research , 2023 , 62 ( 1 ): 519 - 534 .

杜先君 , 邱小彧 . 基于改进LSTM神经网络的化工过程故障诊断 [J ] . 兰州理工大学学报 , 2023 , 49 ( 6 ): 72 - 79 .

Du X J , Qiu X Y . Fault diagnosis for chemical process based on an improved LSTM neural network [J ] . Journal of Lanzhou University of Technology , 2023 , 49 ( 6 ): 72 - 79 .

Zhang Z X , Yang X , Huang J , et al . Layer-wise-residual-driven approach for soft sensing in composite dynamic system based on slow and fast time-varying latent variables [J ] . Chemometrics and Intelligent Laboratory Systems , 2024 , 254 : 105245 .

Yang Z Y , Jia R N , Wang P L , et al . Supervised attention-based bidirectional long short-term memory network for nonlinear dynamic soft sensor application [J ] . ACS Omega , 2023 , 8 ( 4 ): 4196 - 4208 .

张勇 , 赵景波 , 权利敏 . 基于卷积层-注意力机制的长短期记忆网络出水氨氮浓度预测方法 [J ] . 化工学报 , 2024 , 75 ( 12 ): 4679 - 4688 .

Zhang Y , Zhao J B , Quan L M . A prediction method for effluent ammonia nitrogen concentration based on convolutional layer and attention mechanism long short-term memory network [J ] . CIESC Journal , 2024 , 75 ( 12 ): 4679 - 4688 .

Alhajeri M S , Abdullah F , Zhe W , et al . Physics-informed machine learning modeling for predictive control using noisy data [J ] . Chemical Engineering Research and Design , 2022 , 186 : 34 - 49 .

Yu Z H , Zhang Z Y , Jiang Q C , et al . Neural network-based hybrid modeling approach incorporating Bayesian optimization with industrial soft sensor application [J ] . Knowledge-Based Systems , 2024 , 301 : 112341 .

He K M , Zhang X Y , Ren S Q , et al . Deep residual learning for image recognition [C ] // 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) . June 27-30, 2016 , Las Vegas, NV, USA . IEEE , 2016 : 770 - 778 .

Zhang M , Xu B K , Jie J , et al . A novel bidirectional long short-term memory network with weighted attention mechanism for industrial soft sensor development [J ] . IEEE Sensors Journal , 2024 , 24 ( 11 ): 18546 - 18555 .

Zheng X Q , Wu B F , Chen H M , et al . A temporal convolution network-based just-in-time learning method for industrial quality variable prediction [J ] . Chemical Engineering Research and Design , 2024 , 212 : 168 - 184 .

Gao S W , Yun P X , Yang W B , et al . Dynamic soft sensor modelling based on data imputation and spatiotemporal attention [J ] . The Canadian Journal of Chemical Engineering , 2025 , 103 ( 10 ): 4751 - 4773 .

高学金 , 孟令军 , 高慧慧 . 基于注意力LSTM的多阶段发酵过程集成质量预测 [J ] . 控制与决策 , 2022 , 37 ( 3 ): 616 - 624 .

Gao X J , Meng L J , Gao H H . Integrated quality prediction of multi-stage fermentation process based on attention LSTM [J ] . Control and Decision , 2022 , 37 ( 3 ): 616 - 624 .

Queiroz D G , Rodrigues F D B , do Nascimento Pereira Nogueira J , et al . Synergizing phenomenological and AI-based models with industrial data to develop soft sensors for a sour water treatment unit [J ] . Processes , 2024 , 12 ( 9 ): 1900 .

Li S , Jiang Q B , Xu Y D , et al . Digital twin-assisted interpretable transfer learning: A novel wavelet-based framework for intelligent fault diagnostics from simulated domain to real industrial domain [J ] . Advanced Engineering Informatics , 2024 , 62 : 102681 .

Fortuna L , Graziani S , Rizzo A , et al . Soft sensors for monitoring and control of industrial processes [M ] . London : Springer London , 2007 .

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Prediction of three-dimensional temperature fields of data center based on CNN-MLP-ThermoAttention

Application of deep VGG model-based prediction in ethylene cracker plant

Ensemble transfer learning framework for outflow compositions prediction in steam cracking process

Chemical process feature extraction and monitoring method based on latent cointegration autoencoder

Time series data preprocessing in chemical process modeling: Algorithms and applications

Related Author

WANG Fangzhou

LIU Feng

DU Zhongxuan

WU Lili

YUE Yifan

HU Shufan

CAO Jun

ZHU Xiaozhong

Related Institution

School of Mechanical and Power Engineering, East China University of Science and Technology

Shanghai Marine Disel Engine Research Institute

SINOPEC Research Institute of Petroleum Processing

School of Chemical Engineering, Sichuan University

Department of Chemical Engineering, Imperial College London

Postal code：100011
Tel：010-64519466 Email：hgjz@cip.com.cn
Technical support is provided by Beijing Founder electronics co., LTD 京ICP备12046843号-2 京公网安备11010102001994号
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