Computers & Chemical Engineering最新文献

Online alarm flood classification via interpretable template extraction and structured convolutional matching 基于可解释模板提取和结构化卷积匹配的洪水报警在线分类

IF 3.9 2区工程技术 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Chemical Engineering

Pub Date : 2026-05-01 Epub Date: 2026-01-19 DOI: 10.1016/j.compchemeng.2026.109570

Yashar Rahimi , Harikrishna Rao Mohan Rao , Jing Zhou , Tongwen Chen

Alarm flood classification in industrial alarm systems is a challenging task due to variability in fault durations, process noise, and the volume of overlapping alarms. However, alarm floods triggered by similar faults often exhibit recurring structural patterns, which, if identified effectively, can support the root cause diagnosis and informed decision-making by operators. Existing classification methods often rely on opaque models, extensive retraining, or lack integration with operator-facing tools. Motivated by this practical problem, a unified visual analytics-based methodology for the real-time classification of alarm floods is proposed in this paper. The contributions are threefold: (1) The existing High-Density Alarm Plot (HDAP) is extended into a structured matrix representation to encode alarm activity over time; (2) a 2D convolution-based alignment technique is developed to extract representative templates from historical alarm floods, enabling category-specific pattern generation; and (3) a dynamic matrix representation is introduced to support real-time alarm monitoring, where similarity matching against pre-learned templates facilitates online classification with minimal delay. The proposed method is interpretable, operator-friendly, and seamlessly integrates with existing visual tools. The effectiveness of the proposed method is validated on the Tennessee Eastman Process benchmark, demonstrating robust and accurate early-stage classification.

由于故障持续时间、过程噪声和重叠报警量的变化，工业报警系统中的报警洪水分类是一项具有挑战性的任务。然而，由类似故障触发的报警洪水通常表现出重复的结构模式，如果有效识别，可以支持根本原因诊断和运营商的明智决策。现有的分类方法通常依赖于不透明的模型，大量的再培训，或者缺乏与面向操作人员的工具的集成。针对这一实际问题，本文提出了一种统一的基于可视化分析的洪水报警实时分类方法。贡献有三个方面：(1)将现有的高密度报警图（HDAP）扩展为结构化矩阵表示，以编码随时间变化的报警活动；(2)开发了一种基于二维卷积的对齐技术，从历史报警洪水中提取代表性模板，实现了特定类别模式的生成；(3)引入动态矩阵表示来支持实时报警监控，其中针对预学习模板的相似度匹配有助于以最小的延迟在线分类。该方法具有可解释性、操作友好性和与现有可视化工具无缝集成的特点。在田纳西州伊士曼过程基准上验证了该方法的有效性，证明了早期分类的鲁棒性和准确性。

{"title":"Online alarm flood classification via interpretable template extraction and structured convolutional matching","authors":"Yashar Rahimi , Harikrishna Rao Mohan Rao , Jing Zhou , Tongwen Chen","doi":"10.1016/j.compchemeng.2026.109570","DOIUrl":"10.1016/j.compchemeng.2026.109570","url":null,"abstract":"<div><div>Alarm flood classification in industrial alarm systems is a challenging task due to variability in fault durations, process noise, and the volume of overlapping alarms. However, alarm floods triggered by similar faults often exhibit recurring structural patterns, which, if identified effectively, can support the root cause diagnosis and informed decision-making by operators. Existing classification methods often rely on opaque models, extensive retraining, or lack integration with operator-facing tools. Motivated by this practical problem, a unified visual analytics-based methodology for the real-time classification of alarm floods is proposed in this paper. The contributions are threefold: (1) The existing High-Density Alarm Plot (HDAP) is extended into a structured matrix representation to encode alarm activity over time; (2) a 2D convolution-based alignment technique is developed to extract representative templates from historical alarm floods, enabling category-specific pattern generation; and (3) a dynamic matrix representation is introduced to support real-time alarm monitoring, where similarity matching against pre-learned templates facilitates online classification with minimal delay. The proposed method is interpretable, operator-friendly, and seamlessly integrates with existing visual tools. The effectiveness of the proposed method is validated on the Tennessee Eastman Process benchmark, demonstrating robust and accurate early-stage classification.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"208 ","pages":"Article 109570"},"PeriodicalIF":3.9,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146025889","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Transforming electric vehicle battery recycling network with integrated winner determination for enhanced resilience 基于综合赢家判定的电动汽车电池回收网络改造

IF 3.9 2区工程技术 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Chemical Engineering

Pub Date : 2026-05-01 Epub Date: 2026-01-29 DOI: 10.1016/j.compchemeng.2026.109561

Xuefeng Wang , Qiang Liu , Mingqiang Yin

With the rapid development of the electric vehicle industry, the recycling of retired power batteries has become a critical component for enhancing resource efficiency and supporting low-carbon development. To address supply chain disruption risks caused by electric vehicle battery recycling network design (EVBRND), an optimization framework integrating EVBRND and winner determination (WD) is developed from a fourth-party logistics perspective. This study innovatively integrates the WD approach into EVBRND to achieve a cost-effective balance between proprietary recycling networks and external recyclers under disruption risks. A two-stage stochastic resilient model, incorporating a multi-level backup strategy, is formulated and transformed into a mixed-integer linear programming model. To address the computational complexity caused by a large number of disruption scenarios, a scenario-reduction-based decomposition algorithm is proposed. The scenario reduction algorithm is employed to reduce the number of disruption scenarios and to efficiently obtain an upper bound of the optimal solution of the proposed model. Variable relaxation and Lagrangian relaxation algorithms are used to obtain a lower bound of the optimal solution of the proposed model. To verify the effectiveness of the proposed model and algorithms, a real-world case study is conducted. Numerical analysis results indicate that integrating the WD approach improves recycling network responsiveness and economic efficiency.

随着电动汽车产业的快速发展，退役动力电池的回收利用已成为提高资源效率、支持低碳发展的重要组成部分。为了解决电动汽车电池回收网络设计（EVBRND）带来的供应链中断风险，从第四方物流的角度出发，构建了一个整合EVBRND和赢家决策（WD）的优化框架。本研究创新地将WD方法整合到EVBRND中，以实现在中断风险下专有回收网络和外部回收商之间的成本效益平衡。建立了包含多级备份策略的两阶段随机弹性模型，并将其转化为混合整数线性规划模型。针对大量中断场景导致的计算复杂度，提出了一种基于场景约简的分解算法。采用场景约简算法来减少中断场景的数量，并有效地得到模型最优解的上界。采用变松弛算法和拉格朗日松弛算法求出模型最优解的下界。为了验证所提出的模型和算法的有效性，进行了实际案例研究。数值分析结果表明，整合WD方法提高了回收网络的响应能力和经济效率。

{"title":"Transforming electric vehicle battery recycling network with integrated winner determination for enhanced resilience","authors":"Xuefeng Wang , Qiang Liu , Mingqiang Yin","doi":"10.1016/j.compchemeng.2026.109561","DOIUrl":"10.1016/j.compchemeng.2026.109561","url":null,"abstract":"<div><div>With the rapid development of the electric vehicle industry, the recycling of retired power batteries has become a critical component for enhancing resource efficiency and supporting low-carbon development. To address supply chain disruption risks caused by electric vehicle battery recycling network design (EVBRND), an optimization framework integrating EVBRND and winner determination (WD) is developed from a fourth-party logistics perspective. This study innovatively integrates the WD approach into EVBRND to achieve a cost-effective balance between proprietary recycling networks and external recyclers under disruption risks. A two-stage stochastic resilient model, incorporating a multi-level backup strategy, is formulated and transformed into a mixed-integer linear programming model. To address the computational complexity caused by a large number of disruption scenarios, a scenario-reduction-based decomposition algorithm is proposed. The scenario reduction algorithm is employed to reduce the number of disruption scenarios and to efficiently obtain an upper bound of the optimal solution of the proposed model. Variable relaxation and Lagrangian relaxation algorithms are used to obtain a lower bound of the optimal solution of the proposed model. To verify the effectiveness of the proposed model and algorithms, a real-world case study is conducted. Numerical analysis results indicate that integrating the WD approach improves recycling network responsiveness and economic efficiency.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"208 ","pages":"Article 109561"},"PeriodicalIF":3.9,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146170557","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Partial least-squares model adaptation by bootstrap resampling 自举重采样自适应偏最小二乘模型

IF 3.9 2区工程技术 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Chemical Engineering

Pub Date : 2026-05-01 Epub Date: 2026-02-02 DOI: 10.1016/j.compchemeng.2026.109585

Elia Arnese-Feffin , Jinwook Rhyu , Benjamin T. Smith , Chris D. Castro , Jacqueline M. Wolfrum , Stacy L. Springs , Roger A. Hart , Tom Mistretta , Richard D. Braatz

Soft sensors play a crucial role in biomanufacturing, the prime example being the estimation of product quality attributes using only easy-to-measure variables from the plant instrumentation. Data-driven models, such as partial least-squares regression, are widely used to this end. However, their predictive performance might degrade whenever process conditions shift, even after changes made on purpose by operators. Model adaptation strategies can keep the model up to date, but they tend to under-deliver when the process undergoes abrupt changes. In this study, we propose a novel model adaptation method that exploits knowledge of intentional abrupt changes in the process via bootstrap resampling. Exploiting such knowledge, the available data can be used in an optimal way. We demonstrate our approach on three case studies: a numerical example, a simulated penicillin production process, and an industrial biomanufacturing process. Compared to existing adaptation strategies, our method achieves faster adaptation and better predictive performance in the transition period between old and new process conditions. The proposed approach thus enables practitioners to quickly recover the predictive power of soft sensors, with clear benefits on process operation and product quality.

软传感器在生物制造中起着至关重要的作用，主要的例子是仅使用来自工厂仪器的易于测量的变量来估计产品质量属性。数据驱动模型，如偏最小二乘回归，被广泛用于这一目的。然而，每当工艺条件发生变化时，即使是在运营商有意进行更改之后，它们的预测性能也可能会下降。模型适应策略可以使模型保持最新，但是当过程经历突然变化时，它们往往交付不足。在本研究中，我们提出了一种新的模型自适应方法，该方法通过自举重采样来利用过程中有意突变的知识。利用这些知识，可以以最优的方式使用可用的数据。我们在三个案例研究中展示了我们的方法：一个数值例子，一个模拟青霉素生产过程，和一个工业生物制造过程。与现有的自适应策略相比，该方法在新旧工艺条件之间的过渡时期实现了更快的自适应和更好的预测性能。因此，所提出的方法使从业者能够快速恢复软传感器的预测能力，对过程操作和产品质量有明显的好处。

{"title":"Partial least-squares model adaptation by bootstrap resampling","authors":"Elia Arnese-Feffin , Jinwook Rhyu , Benjamin T. Smith , Chris D. Castro , Jacqueline M. Wolfrum , Stacy L. Springs , Roger A. Hart , Tom Mistretta , Richard D. Braatz","doi":"10.1016/j.compchemeng.2026.109585","DOIUrl":"10.1016/j.compchemeng.2026.109585","url":null,"abstract":"<div><div>Soft sensors play a crucial role in biomanufacturing, the prime example being the estimation of product quality attributes using only easy-to-measure variables from the plant instrumentation. Data-driven models, such as partial least-squares regression, are widely used to this end. However, their predictive performance might degrade whenever process conditions shift, even after changes made on purpose by operators. Model adaptation strategies can keep the model up to date, but they tend to under-deliver when the process undergoes abrupt changes. In this study, we propose a novel model adaptation method that exploits knowledge of intentional abrupt changes in the process via bootstrap resampling. Exploiting such knowledge, the available data can be used in an optimal way. We demonstrate our approach on three case studies: a numerical example, a simulated penicillin production process, and an industrial biomanufacturing process. Compared to existing adaptation strategies, our method achieves faster adaptation and better predictive performance in the transition period between old and new process conditions. The proposed approach thus enables practitioners to quickly recover the predictive power of soft sensors, with clear benefits on process operation and product quality.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"208 ","pages":"Article 109585"},"PeriodicalIF":3.9,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146170560","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Modelling and optimization of extractive distillation for IPA-water separation using artificial neural network models 基于人工神经网络模型的ipa -水萃取精馏建模与优化

IF 3.9 2区工程技术 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Chemical Engineering

Pub Date : 2026-05-01 Epub Date: 2026-01-22 DOI: 10.1016/j.compchemeng.2026.109580

Ziba Valizadeh, Hanieh Shokrkar

The extraction of isopropyl alcohol (IPA) from water has long presented a considerable challenge in the chemical industry, mainly due to the formation of an azeotropic mixture and the close boiling points of the two substances. This research introduces a hybrid extractive distillation method utilizing dimethyl sulfoxide (DMSO) as an entrainer, which was simulated using Aspen Plus V11 (NRTL model). The comprehensive simulation attained an impressive IPA purity of 99.9%, which resulted in considerable energy costs. In order to tackle the balance between purity and energy consumption, a novel hybrid framework was created, and essential operational parameters such as feed stage, reflux ratio, an entrainer ratio, condenser duty, and reboiler duty were fine-tuned to enhance IPA purity while reducing energy consumption. The advancement is found in substituting conventional simulation methods with a trained artificial neural network (ANN) to enable rapid predictions, alongside the use of particle swarm optimization (PSO) for fine-tuning parameters. The ANN-PSO framework successfully pinpointed an optimal operating point that led to a 25% decrease in total energy consumption when compared to the baseline. While achieving a practically acceptable IPA purity of 95% in the first column and a water purity of 95% in the second column at reflux ratios of 0.77 and 0.4, respectively, this method demonstrates a notable decrease in computational effort (40% less time) and offers a reliable strategy for low-energy separation processes at an industrial level.

长期以来，从水中提取异丙醇（IPA）一直是化学工业面临的一个相当大的挑战，主要是由于形成共沸混合物和两种物质的沸点接近。本文介绍了一种以二甲亚砜（DMSO）为夹带剂的混合萃取精馏方法，并采用Aspen Plus V11 （NRTL模型）进行了模拟。综合模拟获得了令人印象深刻的99.9%的IPA纯度，这导致了相当大的能源成本。为了解决纯度和能耗之间的平衡，创建了一个新的混合框架，并对进料级、回流比、夹带器比、冷凝器负荷和再锅炉负荷等基本操作参数进行了调整，以提高IPA纯度，同时降低能耗。这一进步是用训练有素的人工神经网络（ANN）代替传统的模拟方法来实现快速预测，同时使用粒子群优化（PSO）来微调参数。ANN-PSO框架成功地确定了一个最佳工作点，与基线相比，总能耗降低了25%。虽然在回流比分别为0.77和0.4的情况下，第一柱的IPA纯度为95%，第二柱的水纯度为95%，但该方法可以显著减少计算工作量（减少40%的时间），并为工业水平的低能量分离过程提供了可靠的策略。

{"title":"Modelling and optimization of extractive distillation for IPA-water separation using artificial neural network models","authors":"Ziba Valizadeh, Hanieh Shokrkar","doi":"10.1016/j.compchemeng.2026.109580","DOIUrl":"10.1016/j.compchemeng.2026.109580","url":null,"abstract":"<div><div>The extraction of isopropyl alcohol (IPA) from water has long presented a considerable challenge in the chemical industry, mainly due to the formation of an azeotropic mixture and the close boiling points of the two substances. This research introduces a hybrid extractive distillation method utilizing dimethyl sulfoxide (DMSO) as an entrainer, which was simulated using Aspen Plus V11 (NRTL model). The comprehensive simulation attained an impressive IPA purity of 99.9%, which resulted in considerable energy costs. In order to tackle the balance between purity and energy consumption, a novel hybrid framework was created, and essential operational parameters such as feed stage, reflux ratio, an entrainer ratio, condenser duty, and reboiler duty were fine-tuned to enhance IPA purity while reducing energy consumption. The advancement is found in substituting conventional simulation methods with a trained artificial neural network (ANN) to enable rapid predictions, alongside the use of particle swarm optimization (PSO) for fine-tuning parameters. The ANN-PSO framework successfully pinpointed an optimal operating point that led to a 25% decrease in total energy consumption when compared to the baseline. While achieving a practically acceptable IPA purity of 95% in the first column and a water purity of 95% in the second column at reflux ratios of 0.77 and 0.4, respectively, this method demonstrates a notable decrease in computational effort (40% less time) and offers a reliable strategy for low-energy separation processes at an industrial level.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"208 ","pages":"Article 109580"},"PeriodicalIF":3.9,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146075820","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

YANNs: Y-wise affine neural networks for exact and efficient representations of piecewise linear functions YANNs：用于分段线性函数的精确和有效表示的y向仿射神经网络

IF 3.9 2区工程技术 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Chemical Engineering

Pub Date : 2026-05-01 Epub Date: 2026-02-10 DOI: 10.1016/j.compchemeng.2026.109589

Austin Braniff, Yuhe Tian

This work formally introduces Y-wise Affine Neural Networks (YANNs), a fully-explainable network architecture that continuously and efficiently represent piecewise affine functions with polytopic subdomains. Following from the proofs, it is shown that the development of YANNs requires no training to achieve the functionally equivalent representation. YANNs thus maintain all mathematical properties of the original formulations. Multi-parametric model predictive control is utilized as an application showcase of YANNs, which theoretically computes optimal control laws as a piecewise affine function of states, outputs, setpoints, and disturbances. With the exact representation of multi-parametric control laws, YANNs retain essential control-theoretic guarantees such as recursive feasibility and stability. This sets YANNs apart from the existing works which apply neural networks for approximating optimal control laws instead of exactly representing them. By optimizing the inference speed of the networks, YANNs can evaluate substantially faster in real-time compared to traditional piecewise affine function calculations. Numerical case studies are presented to demonstrate the algorithmic scalability with respect to the input/output dimensions and the number of subdomains. Future applications can leverage them as an efficient and interpretable starting point for data-driven modeling/control.

这项工作正式介绍了Y-wise仿射神经网络（yann），这是一种完全可解释的网络架构，可以连续有效地表示具有多面体子域的分段仿射函数。根据证明，yann的开发不需要训练就可以实现功能等效表示。因此，yann保持了原始公式的所有数学性质。多参数模型预测控制被用作yann的一个应用展示，它从理论上计算最优控制律作为状态、输出、设定点和干扰的分段仿射函数。由于多参数控制律的精确表示，yann保留了递归可行性和稳定性等基本控制理论保证。这将yann与现有的应用神经网络来近似最优控制律而不是精确表示它们的工作区分开来。通过优化网络的推理速度，与传统的分段仿射函数计算相比，yann可以更快地实时评估。通过数值实例研究，证明了算法在输入/输出维度和子域数量方面的可扩展性。未来的应用程序可以利用它们作为数据驱动建模/控制的高效且可解释的起点。

{"title":"YANNs: Y-wise affine neural networks for exact and efficient representations of piecewise linear functions","authors":"Austin Braniff, Yuhe Tian","doi":"10.1016/j.compchemeng.2026.109589","DOIUrl":"10.1016/j.compchemeng.2026.109589","url":null,"abstract":"<div><div>This work formally introduces Y-wise Affine Neural Networks (YANNs), a fully-explainable network architecture that continuously and efficiently represent piecewise affine functions with polytopic subdomains. Following from the proofs, it is shown that the development of YANNs requires no training to achieve the functionally equivalent representation. YANNs thus maintain all mathematical properties of the original formulations. Multi-parametric model predictive control is utilized as an application showcase of YANNs, which theoretically computes optimal control laws as a piecewise affine function of states, outputs, setpoints, and disturbances. With the exact representation of multi-parametric control laws, YANNs retain essential control-theoretic guarantees such as recursive feasibility and stability. This sets YANNs apart from the existing works which apply neural networks for approximating optimal control laws instead of exactly representing them. By optimizing the inference speed of the networks, YANNs can evaluate substantially faster in real-time compared to traditional piecewise affine function calculations. Numerical case studies are presented to demonstrate the algorithmic scalability with respect to the input/output dimensions and the number of subdomains. Future applications can leverage them as an efficient and interpretable starting point for data-driven modeling/control.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"208 ","pages":"Article 109589"},"PeriodicalIF":3.9,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146170561","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Bayesian optimization and temporal attention-enhanced deep neural network for accurate and reliable state of health estimation of lithium-ion batteries 基于贝叶斯优化和时间注意力增强的深度神经网络的锂离子电池健康状态准确可靠估计

IF 3.9 2区工程技术 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Chemical Engineering

Pub Date : 2026-05-01 Epub Date: 2026-01-19 DOI: 10.1016/j.compchemeng.2026.109571

Zhiyu Chen , Hanfei Wang , Siquan Li , Kena Chen , Jinjie Wang , Ping Wang , Lijun Yang

The capacity degradation trend can indirectly reflect the health status of a battery, and accurate state of health (SOH) estimation can reduce the risk of failure and ensure stable operation. However, the limited feature learning capability of traditional models and the random combination of hyperparameters often lead to large estimation errors. To address the issues, this research proposes a deep neural network (DNN) enhanced by Bayesian optimization (BO) and the temporal attention (TA) mechanism to achieve accurate and reliable SOH estimation for lithium-ion batteries. First, direct aging features of the battery are extracted based on the constant-current charging phase and further fused using principal component analysis (PCA). Then, a mapping model between the aging features and capacity is constructed, in which the TA mechanism is employed to enhance the feature learning capability of the DNN, and BO is used to determine the optimal combination of key hyperparameters. Finally, twelve single-battery under different operating conditions and seven multi-battery capacity estimation experiments are conducted. The estimation performance of the proposed model is evaluated using metrics such as mean absolute error (MAE). The experimental results show that the BO-TADNN model achieves capacity estimation errors within ±3% for single-battery experiments, representing an improvement of approximately 70% in stability compared to the DNN. Furthermore, BO-TADNN achieves the best performance across all evaluation metrics in the multi-battery experiments, which provides a theoretical foundation for future applications in battery management systems.

容量退化趋势可以间接反映电池的健康状态，准确的健康状态（SOH）估计可以降低电池故障的风险，保证电池的稳定运行。然而，传统模型有限的特征学习能力和超参数的随机组合往往导致估计误差较大。为了解决这一问题，本研究提出了一种基于贝叶斯优化（BO）和时间注意（TA）机制的深度神经网络（DNN）来实现锂离子电池SOH的准确可靠估计。首先，基于恒流充电阶段提取电池的直接老化特征，并利用主成分分析（PCA）进行融合；然后，构建了老化特征与容量的映射模型，其中利用TA机制增强深度神经网络的特征学习能力，利用BO确定关键超参数的最优组合。最后进行了12个不同工况下的单电池和7个多电池容量估算实验。使用平均绝对误差（MAE）等度量来评估所提出模型的估计性能。实验结果表明，BO-TADNN模型在单电池实验中实现了±3%的容量估计误差，与DNN相比，稳定性提高了约70%。此外，在多电池实验中，BO-TADNN在所有评估指标中都取得了最佳性能，为未来在电池管理系统中的应用提供了理论基础。

{"title":"Bayesian optimization and temporal attention-enhanced deep neural network for accurate and reliable state of health estimation of lithium-ion batteries","authors":"Zhiyu Chen , Hanfei Wang , Siquan Li , Kena Chen , Jinjie Wang , Ping Wang , Lijun Yang","doi":"10.1016/j.compchemeng.2026.109571","DOIUrl":"10.1016/j.compchemeng.2026.109571","url":null,"abstract":"<div><div>The capacity degradation trend can indirectly reflect the health status of a battery, and accurate state of health (SOH) estimation can reduce the risk of failure and ensure stable operation. However, the limited feature learning capability of traditional models and the random combination of hyperparameters often lead to large estimation errors. To address the issues, this research proposes a deep neural network (DNN) enhanced by Bayesian optimization (BO) and the temporal attention (TA) mechanism to achieve accurate and reliable SOH estimation for lithium-ion batteries. First, direct aging features of the battery are extracted based on the constant-current charging phase and further fused using principal component analysis (PCA). Then, a mapping model between the aging features and capacity is constructed, in which the TA mechanism is employed to enhance the feature learning capability of the DNN, and BO is used to determine the optimal combination of key hyperparameters. Finally, twelve single-battery under different operating conditions and seven multi-battery capacity estimation experiments are conducted. The estimation performance of the proposed model is evaluated using metrics such as mean absolute error (MAE). The experimental results show that the BO-TADNN model achieves capacity estimation errors within ±3% for single-battery experiments, representing an improvement of approximately 70% in stability compared to the DNN. Furthermore, BO-TADNN achieves the best performance across all evaluation metrics in the multi-battery experiments, which provides a theoretical foundation for future applications in battery management systems.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"208 ","pages":"Article 109571"},"PeriodicalIF":3.9,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146006595","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

PSFCL: A Probabilistic Slow Feature Contrastive Learning approach for incipient fault diagnosis in industrial processes 基于概率慢特征对比学习的工业过程早期故障诊断方法

IF 3.9 2区工程技术 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Chemical Engineering

Pub Date : 2026-05-01 Epub Date: 2026-01-30 DOI: 10.1016/j.compchemeng.2026.109584

Liangliang Shang , Rui Fang , Jie Liu , Yingzi Jing , Wan Chen , Aibing Qiu

Recent advances in machine learning have significantly improved fault diagnosis through refined feature extraction and intelligent classification. However, traditional methods often fail to effectively capture the incipient fault features in complex industrial processes. To address this challenge, this article proposes a novel framework named Probabilistic Slow Feature Contrastive Learning (PSFCL) for enhanced incipient feature representation and fault diagnosis. Unlike conventional approaches, the proposed method integrates divergence to extract probabilistic slow features that encode both temporal slowness and distributional distinctiveness. These features are then used to generate positive and negative sample pairs for contrastive learning, enabling the model to learn more discriminative representations. Furthermore, a slowness-guided mechanism adaptively adjusts model parameters during training, reinforcing fault feature extraction. The learned features are fed into a Softmax classifier for incipient fault diagnosis without requiring parameter fine-tuning. Benchmark-based validation on the Tennessee Eastman Process (TEP) under diverse fault types demonstrates that the proposed PSFCL framework outperforms state-of-the-art methods in both diagnostic accuracy and robustness, thereby showcasing its potential to maintain industrial process safety.

机器学习的最新进展通过精细特征提取和智能分类显著改善了故障诊断。然而，传统的方法往往不能有效地捕捉复杂工业过程的早期故障特征。为了解决这一挑战，本文提出了一种名为概率慢特征对比学习（PSFCL）的新框架，用于增强早期特征表示和故障诊断。与传统方法不同，该方法集成了散度来提取概率慢特征，该特征编码了时间慢性和分布独特性。然后使用这些特征生成正负样本对进行对比学习，使模型能够学习更多的判别表示。此外，在训练过程中，慢速引导机制自适应调整模型参数，加强故障特征提取。将学习到的特征输入到Softmax分类器中进行早期故障诊断，而不需要对参数进行微调。在田纳西伊士曼过程（TEP）不同故障类型下基于基准的验证表明，所提出的PSFCL框架在诊断准确性和鲁棒性方面都优于最先进的方法，从而展示了其维护工业过程安全的潜力。

{"title":"PSFCL: A Probabilistic Slow Feature Contrastive Learning approach for incipient fault diagnosis in industrial processes","authors":"Liangliang Shang , Rui Fang , Jie Liu , Yingzi Jing , Wan Chen , Aibing Qiu","doi":"10.1016/j.compchemeng.2026.109584","DOIUrl":"10.1016/j.compchemeng.2026.109584","url":null,"abstract":"<div><div>Recent advances in machine learning have significantly improved fault diagnosis through refined feature extraction and intelligent classification. However, traditional methods often fail to effectively capture the incipient fault features in complex industrial processes. To address this challenge, this article proposes a novel framework named Probabilistic Slow Feature Contrastive Learning (PSFCL) for enhanced incipient feature representation and fault diagnosis. Unlike conventional approaches, the proposed method integrates divergence to extract probabilistic slow features that encode both temporal slowness and distributional distinctiveness. These features are then used to generate positive and negative sample pairs for contrastive learning, enabling the model to learn more discriminative representations. Furthermore, a slowness-guided mechanism adaptively adjusts model parameters during training, reinforcing fault feature extraction. The learned features are fed into a Softmax classifier for incipient fault diagnosis without requiring parameter fine-tuning. Benchmark-based validation on the Tennessee Eastman Process (TEP) under diverse fault types demonstrates that the proposed PSFCL framework outperforms state-of-the-art methods in both diagnostic accuracy and robustness, thereby showcasing its potential to maintain industrial process safety.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"208 ","pages":"Article 109584"},"PeriodicalIF":3.9,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146170553","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Robust optimal experimental design for identifiability using the overall mean squared estimation error 使用总体均方估计误差的可辨识性稳健优化实验设计

IF 3.9 2区工程技术 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Chemical Engineering

Pub Date : 2026-05-01 Epub Date: 2026-01-24 DOI: 10.1016/j.compchemeng.2026.109581

Pierre Denis , Constantinos Theodoropoulos , Alain Vande Wouwer

Mathematical modeling is essential for understanding and controlling physical systems, particularly in scientific and engineering contexts. Accurate parameter estimation is critical for model reliability but often constrained by the cost and complexity of experiments. Optimal Experimental Design (OED) addresses this challenge by identifying experimental conditions that maximize information gain. Traditional OED approaches rely on the Fisher Information Matrix (FIM) and scalar optimality criteria, yet they are sensitive to unknown parameter values. To mitigate this, robust OED methods incorporate prior uncertainty, using strategies such as maximin and expectation-based criteria. In this work, we introduce a novel robust OED framework that unifies prior parameter uncertainty and measurement noise into an Overall Mean Squared estimation Error (OMSE) matrix. This formulation enables the use of standard optimality criteria while inherently accounting for both sources of uncertainty/noise. We demonstrate the effectiveness of our method through two case studies involving dynamical systems of varying complexity and discuss practical considerations for its implementation.

数学建模对于理解和控制物理系统至关重要，特别是在科学和工程环境中。准确的参数估计对模型可靠性至关重要，但往往受到实验成本和复杂性的限制。优化实验设计（OED）通过确定最大化信息增益的实验条件来解决这一挑战。传统的OED方法依赖于Fisher信息矩阵（FIM）和标量最优性准则，但它们对未知参数值很敏感。为了减轻这种情况，健壮的OED方法使用诸如最大化和基于期望的标准等策略，将先前的不确定性纳入其中。在这项工作中，我们引入了一种新的鲁棒OED框架，该框架将先验参数不确定性和测量噪声统一到总体均方估计误差（OMSE）矩阵中。这个公式可以使用标准的最优性标准，同时内在地考虑到不确定性/噪声的两个来源。我们通过涉及不同复杂性的动态系统的两个案例研究证明了我们方法的有效性，并讨论了其实施的实际考虑因素。

{"title":"Robust optimal experimental design for identifiability using the overall mean squared estimation error","authors":"Pierre Denis , Constantinos Theodoropoulos , Alain Vande Wouwer","doi":"10.1016/j.compchemeng.2026.109581","DOIUrl":"10.1016/j.compchemeng.2026.109581","url":null,"abstract":"<div><div>Mathematical modeling is essential for understanding and controlling physical systems, particularly in scientific and engineering contexts. Accurate parameter estimation is critical for model reliability but often constrained by the cost and complexity of experiments. Optimal Experimental Design (OED) addresses this challenge by identifying experimental conditions that maximize information gain. Traditional OED approaches rely on the Fisher Information Matrix (FIM) and scalar optimality criteria, yet they are sensitive to unknown parameter values. To mitigate this, robust OED methods incorporate prior uncertainty, using strategies such as maximin and expectation-based criteria. In this work, we introduce a novel robust OED framework that unifies prior parameter uncertainty and measurement noise into an Overall Mean Squared estimation Error (OMSE) matrix. This formulation enables the use of standard optimality criteria while inherently accounting for both sources of uncertainty/noise. We demonstrate the effectiveness of our method through two case studies involving dynamical systems of varying complexity and discuss practical considerations for its implementation.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"208 ","pages":"Article 109581"},"PeriodicalIF":3.9,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146170556","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Digital twin framework with physics-informed neural networks for real-time monitoring of PEM electrolyzers in renewable microgrids 用于可再生微电网PEM电解槽实时监测的物理信息神经网络数字孪生框架

IF 3.9 2区工程技术 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Chemical Engineering

Pub Date : 2026-05-01 Epub Date: 2026-01-25 DOI: 10.1016/j.compchemeng.2026.109582

Hassan NAANANI, Meriem KAYSOUNY, Anas ABERHOUCH, Said SAIR, Abdessamad FAIK

The operation of electrolyzers for producing green hydrogen faces two key challenges: the rapidly increasing demand for hydrogen across diverse applications and the limited durability of electrochemical components. Digital twin (DT) technology offers a promising pathway to address these limitations by enabling real-time monitoring, fault detection, and predictive analysis. This study presents the development of a DT for a laboratory-scale proton exchange membrane (PEM) electrolyzer composed of two series-connected cells. The virtual counterpart synchronizes with experimental voltage and current data acquired under varying operating conditions, enabling continuous verification of the system’s behavior. To enhance predictive capability, a physicsinformed neural networks (PINN) is integrated into the DT, combining polarization data with electrochemical constraints, including Butler-Volmer activation kinetics and ohmic resistance. A complementary 3D representation of the developed system provides an interactive visualization of the electrolyzer and its operating state. The resulting framework supports real-time supervision, performance assessment, and degradation monitoring, offering a practical foundation for intelligent control and diagnostic strategies in PEM electrolysis systems.

生产绿色氢的电解槽的运行面临两个关键挑战：各种应用对氢的需求快速增长以及电化学元件的有限耐用性。数字孪生（DT）技术通过实现实时监控、故障检测和预测分析，为解决这些限制提供了一条很有前途的途径。本研究介绍了一个由两个串联电池组成的实验室规模质子交换膜（PEM）电解槽的DT的发展。虚拟对应体与在不同操作条件下获得的实验电压和电流数据同步，从而能够连续验证系统的行为。为了提高预测能力，将物理信息神经网络（PINN）集成到DT中，将极化数据与电化学约束（包括Butler-Volmer活化动力学和欧姆电阻）相结合。所开发系统的补充3D表示提供了电解槽及其运行状态的交互式可视化。由此产生的框架支持实时监督、性能评估和退化监测，为PEM电解系统的智能控制和诊断策略提供了实用基础。

{"title":"Digital twin framework with physics-informed neural networks for real-time monitoring of PEM electrolyzers in renewable microgrids","authors":"Hassan NAANANI, Meriem KAYSOUNY, Anas ABERHOUCH, Said SAIR, Abdessamad FAIK","doi":"10.1016/j.compchemeng.2026.109582","DOIUrl":"10.1016/j.compchemeng.2026.109582","url":null,"abstract":"<div><div>The operation of electrolyzers for producing green hydrogen faces two key challenges: the rapidly increasing demand for hydrogen across diverse applications and the limited durability of electrochemical components. Digital twin (DT) technology offers a promising pathway to address these limitations by enabling real-time monitoring, fault detection, and predictive analysis. This study presents the development of a DT for a laboratory-scale proton exchange membrane (PEM) electrolyzer composed of two series-connected cells. The virtual counterpart synchronizes with experimental voltage and current data acquired under varying operating conditions, enabling continuous verification of the system’s behavior. To enhance predictive capability, a physicsinformed neural networks (PINN) is integrated into the DT, combining polarization data with electrochemical constraints, including Butler-Volmer activation kinetics and ohmic resistance. A complementary 3D representation of the developed system provides an interactive visualization of the electrolyzer and its operating state. The resulting framework supports real-time supervision, performance assessment, and degradation monitoring, offering a practical foundation for intelligent control and diagnostic strategies in PEM electrolysis systems.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"208 ","pages":"Article 109582"},"PeriodicalIF":3.9,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146075819","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Modeling pH gradients in industrial-scale lactic acid bacteria fermentation using flow-informed compartment models 利用流动信息室模型模拟工业规模乳酸菌发酵中的pH梯度

IF 3.9 2区工程技术 Q2 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Chemical Engineering

Pub Date : 2026-05-01 Epub Date: 2026-01-23 DOI: 10.1016/j.compchemeng.2026.109568

Johan Le Nepvou De Carfort , Michail Raptotasios , Víctor Puig-I-Laborda , Krist V. Gernaey , Ulrich Krühne

This study presents a numerical model of an industrial-scale lactic acid bacteria fermentation process that incorporates reactor hydrodynamics and chemical/biochemical reactions to resolve the impact of transport limitations on the industrial process. The model employs a three-dimensional CFD-based compartmentalization method that uses unsupervised clustering to build a simplified representation of the reactor volume while preserving essential mixing characteristics. This approach enabled fast simulations of the mixing dynamics over extended periods and facilitated the integration with reaction kinetics. The numerical model successfully simulated both batch and continuous processes, capturing the magnitude of gradients and quantifying their impact on process dynamics. During batch operation, pH gradients peaked during exponential growth but did not significantly affect biological growth, although more sensitive strains could be impacted. Continuous operation resulted in gradients in pH and substrate concentration, affecting the process dynamics and productivity. The compartment model’s adaptability to different reactor designs and scales makes it a valuable tool for industrial applications. Here, the model can be used to develop effective scale-up strategies and ensure consistent product quality across the various production scales.

本研究提出了一个工业规模乳酸菌发酵过程的数值模型，该模型结合了反应器流体动力学和化学/生化反应，以解决运输限制对工业过程的影响。该模型采用基于cfd的三维划分方法，该方法使用无监督聚类来构建反应器体积的简化表示，同时保留了基本的混合特征。这种方法能够在较长时间内快速模拟混合动力学，并促进了与反应动力学的整合。数值模型成功地模拟了批量和连续过程，捕获了梯度的大小并量化了它们对过程动力学的影响。在批量操作中，pH梯度在指数生长期间达到峰值，但对生物生长没有显著影响，尽管对更敏感的菌株可能会受到影响。连续操作导致pH和底物浓度的梯度，影响过程动态和生产率。隔间模型对不同反应器设计和规模的适应性使其成为工业应用的宝贵工具。在这里，该模型可用于制定有效的规模扩大策略，并确保不同生产规模的产品质量一致。

{"title":"Modeling pH gradients in industrial-scale lactic acid bacteria fermentation using flow-informed compartment models","authors":"Johan Le Nepvou De Carfort , Michail Raptotasios , Víctor Puig-I-Laborda , Krist V. Gernaey , Ulrich Krühne","doi":"10.1016/j.compchemeng.2026.109568","DOIUrl":"10.1016/j.compchemeng.2026.109568","url":null,"abstract":"<div><div>This study presents a numerical model of an industrial-scale lactic acid bacteria fermentation process that incorporates reactor hydrodynamics and chemical/biochemical reactions to resolve the impact of transport limitations on the industrial process. The model employs a three-dimensional CFD-based compartmentalization method that uses unsupervised clustering to build a simplified representation of the reactor volume while preserving essential mixing characteristics. This approach enabled fast simulations of the mixing dynamics over extended periods and facilitated the integration with reaction kinetics. The numerical model successfully simulated both batch and continuous processes, capturing the magnitude of gradients and quantifying their impact on process dynamics. During batch operation, pH gradients peaked during exponential growth but did not significantly affect biological growth, although more sensitive strains could be impacted. Continuous operation resulted in gradients in pH and substrate concentration, affecting the process dynamics and productivity. The compartment model’s adaptability to different reactor designs and scales makes it a valuable tool for industrial applications. Here, the model can be used to develop effective scale-up strategies and ensure consistent product quality across the various production scales.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"208 ","pages":"Article 109568"},"PeriodicalIF":3.9,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146170555","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0