Computers & Chemical Engineering最新文献_第10页

Data-driven contextual robust optimization based on support vector clustering 基于支持向量聚类的数据驱动上下文鲁棒优化

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

Computers & Chemical Engineering

Pub Date : 2025-01-17 DOI: 10.1016/j.compchemeng.2025.109004

Xianyu Li , Fenglian Dong , Zhiwei Wei , Chao Shang

Support vector clustering (SVC) is an effective data-driven method to construct uncertainty sets in robust optimization (RO). However, it cannot appropriately address varying uncertainty in a contextually uncertain environment. In this work, we propose a new contextual RO (CRO) scheme, where an efficient contextual uncertainty set called kNN-SVC is developed to capture the correlation between covariates and uncertainty. Using the k-nearest neighbors (kNN) to select a subset of historical observations, contextual information can be integrated into SVC uncertainty sets, thereby alleviating conservatism while inheriting merits of SVC such as polytopic representability and ease of manipulating robustness. Besides, using only a fraction of data samples ensures low computational costs. Numerical examples demonstrate the performance improvement of the proposed kNN-SVC uncertainty set over conventional sets without considering contextual information. An industrial case of gasoline blending shows the usefulness of the proposed approach in producing robust decisions against linearization errors in nonlinear blending.

支持向量聚类（SVC）是鲁棒优化中构造不确定性集的有效数据驱动方法。然而，它不能适当地处理上下文不确定环境中的各种不确定性。在这项工作中，我们提出了一种新的上下文RO （CRO）方案，其中开发了一个称为kNN-SVC的高效上下文不确定性集来捕获协变量与不确定性之间的相关性。使用k近邻（kNN）选择历史观测的子集，上下文信息可以集成到SVC的不确定性集中，从而在继承SVC的优点（如多向可表示性和易于操作的鲁棒性）的同时减轻保守性。此外，只使用一小部分数据样本可以确保较低的计算成本。数值算例表明，在不考虑上下文信息的情况下，所提出的kNN-SVC不确定性集的性能优于传统的不确定性集。一个汽油混合的工业实例表明，该方法在非线性混合中对线性化误差产生鲁棒决策方面是有用的。

{"title":"Data-driven contextual robust optimization based on support vector clustering","authors":"Xianyu Li , Fenglian Dong , Zhiwei Wei , Chao Shang","doi":"10.1016/j.compchemeng.2025.109004","DOIUrl":"10.1016/j.compchemeng.2025.109004","url":null,"abstract":"<div><div>Support vector clustering (SVC) is an effective data-driven method to construct uncertainty sets in robust optimization (RO). However, it cannot appropriately address varying uncertainty in a contextually uncertain environment. In this work, we propose a new contextual RO (CRO) scheme, where an efficient contextual uncertainty set called kNN-SVC is developed to capture the correlation between covariates and uncertainty. Using the k-nearest neighbors (kNN) to select a subset of historical observations, contextual information can be integrated into SVC uncertainty sets, thereby alleviating conservatism while inheriting merits of SVC such as polytopic representability and ease of manipulating robustness. Besides, using only a fraction of data samples ensures low computational costs. Numerical examples demonstrate the performance improvement of the proposed kNN-SVC uncertainty set over conventional sets without considering contextual information. An industrial case of gasoline blending shows the usefulness of the proposed approach in producing robust decisions against linearization errors in nonlinear blending.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"195 ","pages":"Article 109004"},"PeriodicalIF":3.9,"publicationDate":"2025-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143348529","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

Logic-Based Discrete-Steepest Descent: A solution method for process synthesis Generalized Disjunctive Programs 基于逻辑的离散最陡下降：过程综合广义析取规划的一种求解方法

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

Computers & Chemical Engineering

Pub Date : 2025-01-17 DOI: 10.1016/j.compchemeng.2024.108993

Daniel Ovalle , David A. Liñán , Albert Lee , Jorge M. Gómez , Luis Ricardez-Sandoval , Ignacio E. Grossmann , David E. Bernal Neira

Optimization of chemical processes is challenging due to nonlinearities arising from chemical principles and discrete design decisions. The optimal synthesis and design of chemical processes can be posed as a Generalized Disjunctive Programming (GDP) problem. While reformulating GDP problems as Mixed-Integer Nonlinear Programming (MINLP) problems is common, specialized algorithms for GDP remain scarce. This study introduces the Logic-Based Discrete-Steepest Descent Algorithm (LD-SDA) as a solution method for GDP problems involving ordered Boolean variables. LD-SDA transforms these variables into external integer decisions and uses a two-level decomposition: the upper-level sets external configurations, and the lower-level solves the remaining variables, efficiently exploiting the GDP structure. In the case studies presented in this work, including batch processing, reactor superstructures, and distillation columns, LD-SDA consistently outperforms conventional GDP and MINLP solvers, especially as the problem size grows. LD-SDA also proves superior when solving challenging problems where other solvers encounter difficulties finding optimal solutions.

由于化学原理和离散设计决策引起的非线性，化学过程的优化具有挑战性。化工过程的优化合成与设计可归结为广义析取规划问题。虽然将GDP问题重新表述为混合整数非线性规划（MINLP）问题是常见的，但专门的GDP算法仍然很少。本文介绍了基于逻辑的离散最速下降算法（LD-SDA）作为一种涉及有序布尔变量的GDP问题的求解方法。LD-SDA将这些变量转换为外部整数决策，并采用两层分解：上层设置外部配置，下层求解剩余变量，有效地利用了GDP结构。在本工作中提出的案例研究中，包括批处理、反应器上层结构和精馏塔，LD-SDA始终优于传统的GDP和MINLP解决方案，特别是当问题规模增长时。LD-SDA在解决其他求解者难以找到最优解的挑战性问题时也证明了其优越性。

{"title":"Logic-Based Discrete-Steepest Descent: A solution method for process synthesis Generalized Disjunctive Programs","authors":"Daniel Ovalle , David A. Liñán , Albert Lee , Jorge M. Gómez , Luis Ricardez-Sandoval , Ignacio E. Grossmann , David E. Bernal Neira","doi":"10.1016/j.compchemeng.2024.108993","DOIUrl":"10.1016/j.compchemeng.2024.108993","url":null,"abstract":"<div><div>Optimization of chemical processes is challenging due to nonlinearities arising from chemical principles and discrete design decisions. The optimal synthesis and design of chemical processes can be posed as a Generalized Disjunctive Programming (GDP) problem. While reformulating GDP problems as Mixed-Integer Nonlinear Programming (MINLP) problems is common, specialized algorithms for GDP remain scarce. This study introduces the Logic-Based Discrete-Steepest Descent Algorithm (LD-SDA) as a solution method for GDP problems involving ordered Boolean variables. LD-SDA transforms these variables into external integer decisions and uses a two-level decomposition: the upper-level sets external configurations, and the lower-level solves the remaining variables, efficiently exploiting the GDP structure. In the case studies presented in this work, including batch processing, reactor superstructures, and distillation columns, LD-SDA consistently outperforms conventional GDP and MINLP solvers, especially as the problem size grows. LD-SDA also proves superior when solving challenging problems where other solvers encounter difficulties finding optimal solutions.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"195 ","pages":"Article 108993"},"PeriodicalIF":3.9,"publicationDate":"2025-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143348531","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

A detailed MILP model and an ad hoc decomposition algorithm for the operational optimization of gas transport networks 一个详细的MILP模型和一种气体输送网络运行优化的自组织分解算法

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

Computers & Chemical Engineering

Pub Date : 2025-01-17 DOI: 10.1016/j.compchemeng.2025.109006

Lavinia Marina Paola Ghilardi , Francesco Casella , Daniele Barbati , Roberto Palazzo , Emanuele Martelli

Nowadays, the management of natural gas networks primarily relies on the expertise gained by operators over the years. Nevertheless, the need to reduce energy consumption and the progressive installation of electric compressors call for the adoption of systematic optimization tools. This study proposes a Mixed Integer Linear Programming (MILP) model for optimizing the operation of complex real-world gas networks to minimize the environmental impact of the compression work in presence of both gas-turbine driven and electric compressors. The operational problem includes the gas transport dynamic equations, detailed modeling of compressor stations and control valves, while handling complex branch and looped networks with possible reverse flow. To address large-scale problems, a graph reduction procedure and a novel bilevel decomposition algorithm are developed. This methodology, validated with real data, enables the optimization of the nationwide Italian network, comprising 51 compressors and 9727 km of pipes.

如今，天然气网络的管理主要依赖于运营商多年来积累的专业知识。然而，降低能耗的需要和电动压缩机的逐步安装要求采用系统优化工具。本研究提出了一个混合整数线性规划（MILP）模型，用于优化复杂的现实世界气体网络的运行，以最大限度地减少燃气轮机驱动和电动压缩机存在的压缩工作对环境的影响。操作问题包括气体输送动力学方程，压缩机站和控制阀的详细建模，同时处理可能存在逆流的复杂分支和环状网络。为了解决大规模问题，提出了一种新的图约简算法和一种新的二层分解算法。该方法经过了实际数据的验证，能够优化意大利全国范围内的管网，包括51台压缩机和9727公里的管道。

{"title":"A detailed MILP model and an ad hoc decomposition algorithm for the operational optimization of gas transport networks","authors":"Lavinia Marina Paola Ghilardi , Francesco Casella , Daniele Barbati , Roberto Palazzo , Emanuele Martelli","doi":"10.1016/j.compchemeng.2025.109006","DOIUrl":"10.1016/j.compchemeng.2025.109006","url":null,"abstract":"<div><div>Nowadays, the management of natural gas networks primarily relies on the expertise gained by operators over the years. Nevertheless, the need to reduce energy consumption and the progressive installation of electric compressors call for the adoption of systematic optimization tools. This study proposes a Mixed Integer Linear Programming (MILP) model for optimizing the operation of complex real-world gas networks to minimize the environmental impact of the compression work in presence of both gas-turbine driven and electric compressors. The operational problem includes the gas transport dynamic equations, detailed modeling of compressor stations and control valves, while handling complex branch and looped networks with possible reverse flow. To address large-scale problems, a graph reduction procedure and a novel bilevel decomposition algorithm are developed. This methodology, validated with real data, enables the optimization of the nationwide Italian network, comprising 51 compressors and 9727 km of pipes.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"195 ","pages":"Article 109006"},"PeriodicalIF":3.9,"publicationDate":"2025-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143348535","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Optimal sensor placement for contamination detection and identification in water distribution networks under demand uncertainty 需求不确定条件下配水管网污染检测与识别的传感器优化配置

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

Computers & Chemical Engineering

Pub Date : 2025-01-13 DOI: 10.1016/j.compchemeng.2025.109003

Venkata Reddy Palleti

Water Distribution Networks (WDNs) are one of the most important critical infrastructures in any nation. WDNs are often prone to either accidental or intentional contamination. Intentional contamination, like terrorist attacks on WDNs, can lead to poisoned water, causing many fatalities and large economic consequences. In order to protect against these attacks, an efficient sensor network design is required by placing a limited number of sensors in the network. In this work, we will design sensor networks to satisfy two criteria, namely, observability (ability to detect the contamination) and identifiability ability to detect and identify the contamination source). Hydraulic simulations are performed on a WDN subjected to variable demand conditions. We will map the problem of the sensor network to a minimum set cover problem. A greedy heuristic algorithm is used to obtain the sensor network design under variable demand conditions. The proposed methodology is illustrated on a real life WDN.

供水管网是任何国家最重要的基础设施之一。wdn通常容易受到意外或故意的污染。故意污染，如恐怖分子对wdn的袭击，可能导致水中毒，造成许多死亡和巨大的经济后果。为了防止这些攻击，需要通过在网络中放置有限数量的传感器来设计有效的传感器网络。在这项工作中，我们将设计传感器网络以满足两个标准，即可观察性（检测污染的能力）和可识别性（检测和识别污染源的能力）。在可变需求条件下对WDN进行了水力仿真。我们将传感器网络问题映射为最小集覆盖问题。采用贪婪启发式算法求解变需求条件下的传感器网络设计。所提出的方法在一个实际的WDN上进行了说明。

引用次数: 0

Process design for plastic waste pyrolysis: Yield analysis and economic assessment 塑料废弃物热解工艺设计：产率分析与经济评价

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

Computers & Chemical Engineering

Pub Date : 2025-01-11 DOI: 10.1016/j.compchemeng.2025.109001

Hyojin Jung, Yuchan Ahn

The widespread use of plastic products and inadequate waste management systems have rendered plastic waste a major source of environmental pollution, thereby necessitating effective plastic recycling methods. In this study, we propose two pyrolysis processes for recycling plastic waste using Aspen Plus: a base process and an alternative process. The main difference lies in the condensation and separation methods used to produce pyrolysis oil. In the base process, condensation is carried out using a cooler, followed by decomposition with a flash drum. In contrast, the alternative process utilizes a distillation column for both condensation and separation. The results showed that PS achieved the highest oil yield of 81.3 %, while PS/PE exhibited the greatest improvement, with a 27.34 % increase in oil production yield in the alternative process compared to the base process. However, the alternative process incurs higher operating and capital costs due to the precise control requirements, particularly for feedstocks like PS/PE/PP/PET, which had the highest energy production cost at 31.07×10 ⁻⁴ $/ MJ compared to 27.85×10 ⁻⁴ $/ MJ in the base process. Despite these higher costs, the alternative process significantly improved oil production, especially for plastics such as PS and PS/PE. These findings underscore the importance of selecting pyrolysis processes based on feedstock composition and specific recycling goals, highlighting the trade-off between higher yields and increased energy production costs and emphasizing the need for balance.

塑料产品的广泛使用和废物管理系统的不足使塑料废物成为环境污染的主要来源，因此需要有效的塑料回收方法。在本研究中，我们提出了利用Aspen Plus回收塑料废物的两种热解工艺：基础工艺和替代工艺。主要区别在于生产热解油所用的冷凝和分离方法。在基础工艺中，使用冷却器进行冷凝，然后使用闪蒸鼓进行分解。相反，替代工艺利用精馏塔进行冷凝和分离。结果表明，PS工艺的产油率最高，达到81.3%，而PS/PE工艺的产油率提高幅度最大，与基础工艺相比，PS/PE工艺的产油率提高了27.34%。然而，由于精确的控制要求，替代工艺会产生更高的运营和资金成本，特别是对于PS/PE/PP/PET等原料，与基础工艺中的27.85×10 - 4美元/ MJ相比，其能源生产成本最高，为31.07×10−4美元/ MJ。尽管成本较高，但替代工艺显著提高了石油产量，特别是PS和PS/PE等塑料。这些发现强调了根据原料组成和特定回收目标选择热解工艺的重要性，强调了更高的产量和增加的能源生产成本之间的权衡，并强调了平衡的必要性。

{"title":"Process design for plastic waste pyrolysis: Yield analysis and economic assessment","authors":"Hyojin Jung, Yuchan Ahn","doi":"10.1016/j.compchemeng.2025.109001","DOIUrl":"10.1016/j.compchemeng.2025.109001","url":null,"abstract":"<div><div>The widespread use of plastic products and inadequate waste management systems have rendered plastic waste a major source of environmental pollution, thereby necessitating effective plastic recycling methods. In this study, we propose two pyrolysis processes for recycling plastic waste using Aspen Plus: a base process and an alternative process. The main difference lies in the condensation and separation methods used to produce pyrolysis oil. In the base process, condensation is carried out using a cooler, followed by decomposition with a flash drum. In contrast, the alternative process utilizes a distillation column for both condensation and separation. The results showed that PS achieved the highest oil yield of 81.3 %, while PS/PE exhibited the greatest improvement, with a 27.34 % increase in oil production yield in the alternative process compared to the base process. However, the alternative process incurs higher operating and capital costs due to the precise control requirements, particularly for feedstocks like PS/PE/PP/PET, which had the highest energy production cost at 31.07×10 <sup>−4</sup> $/ MJ compared to 27.85×10 <sup>−4</sup> $/ MJ in the base process. Despite these higher costs, the alternative process significantly improved oil production, especially for plastics such as PS and PS/PE. These findings underscore the importance of selecting pyrolysis processes based on feedstock composition and specific recycling goals, highlighting the trade-off between higher yields and increased energy production costs and emphasizing the need for balance.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"194 ","pages":"Article 109001"},"PeriodicalIF":3.9,"publicationDate":"2025-01-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143136738","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

Ensemble transfer learning assisted soft sensor for distributed output inference in chemical processes 集成迁移学习辅助软传感器在化工过程分布式输出推理中的应用

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

Computers & Chemical Engineering

Pub Date : 2025-01-09 DOI: 10.1016/j.compchemeng.2025.109002

Jialiang Zhu, Wangwang Zhu, Yi Liu

Chemical processes with distributed outputs are characterized by various operating conditions, and the scarcity of labeled data poses challenges to the prediction of product quality. An ensemble transfer Gaussian process regression (ETGPR) model is developed for prediction of different quantities of distributed outputs. First, for each test instances from target domain, just-in-time learning is adopted to select distance-based similar instances from source domain in related operating conditions. Mutual information helps create various local models by building diverse input variable sets. Subsequently, Bayesian inference is used to produce the posterior probabilities relative to the test instance, then set as the weights of local prediction. The instance transfer is thus completed via distance-based similar instance selection from source domain for local model construction, and the model performance is improved by the ensemble weighting strategy, concerning the target domain, under diverse operating conditions. Therefore, by utilizing and transferring information from source domain, unsupervised transfer can be implemented with available unlabeled target data. The superiority of ETGPR model is confirmed in the case of modeling the polymerization process with distributed outputs.

具有分布式输出的化学过程具有各种操作条件的特点，标记数据的稀缺性对产品质量的预测提出了挑战。建立了一个集成传递高斯过程回归（ETGPR）模型，用于预测不同数量的分布式输出。首先，对于来自目标域的每个测试实例，采用实时学习方法在相关操作条件下从源域选择基于距离的相似实例。互信息通过构建不同的输入变量集来帮助创建不同的局部模型。然后，使用贝叶斯推理产生相对于测试实例的后验概率，然后将其设置为局部预测的权重。因此，通过从源域选择基于距离的相似实例进行局部模型构建来完成实例转移，并通过在不同操作条件下针对目标域的集成加权策略来提高模型性能。因此，通过对源域信息的利用和传递，可以实现对可用的未标记目标数据的无监督传输。通过对具有分布式输出的聚合过程进行建模，验证了ETGPR模型的优越性。

{"title":"Ensemble transfer learning assisted soft sensor for distributed output inference in chemical processes","authors":"Jialiang Zhu, Wangwang Zhu, Yi Liu","doi":"10.1016/j.compchemeng.2025.109002","DOIUrl":"10.1016/j.compchemeng.2025.109002","url":null,"abstract":"<div><div>Chemical processes with distributed outputs are characterized by various operating conditions, and the scarcity of labeled data poses challenges to the prediction of product quality. An ensemble transfer Gaussian process regression (ETGPR) model is developed for prediction of different quantities of distributed outputs. First, for each test instances from target domain, just-in-time learning is adopted to select distance-based similar instances from source domain in related operating conditions. Mutual information helps create various local models by building diverse input variable sets. Subsequently, Bayesian inference is used to produce the posterior probabilities relative to the test instance, then set as the weights of local prediction. The instance transfer is thus completed via distance-based similar instance selection from source domain for local model construction, and the model performance is improved by the ensemble weighting strategy, concerning the target domain, under diverse operating conditions. Therefore, by utilizing and transferring information from source domain, unsupervised transfer can be implemented with available unlabeled target data. The superiority of ETGPR model is confirmed in the case of modeling the polymerization process with distributed outputs.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"194 ","pages":"Article 109002"},"PeriodicalIF":3.9,"publicationDate":"2025-01-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143136597","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

Multi-objective reaction optimization under uncertainties using expected quantile improvement 基于期望分位数改进的不确定条件下多目标反应优化

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

Computers & Chemical Engineering

Pub Date : 2025-01-09 DOI: 10.1016/j.compchemeng.2024.108983

Jiyizhe Zhang , Daria Semochkina , Naoto Sugisawa , David C. Woods , Alexei A. Lapkin

Multi-objective Bayesian optimization (MOBO) has shown to be a promising tool for reaction development. However, noise is usually inevitable in experimental and chemical processes, and finding reliable solutions is challenging when the noise is unknown or significant. In this study, we focus on finding a set of optimal reaction conditions using multi-objective Euclidian expected quantile improvement (MO-E-EQI) under noisy settings. First, the performance of MO-E-EQI is evaluated by comparing with some recent MOBO algorithms in silico with linear and log-linear heteroscedastic noise structures and different magnitudes. It is noticed that high noise can degrade the performance of MOBO algorithms. MO-E-EQI shows robust performance in terms of hypervolume-based metric, coverage metric and number of solutions on the Pareto front. Finally, MO-E-EQI is implemented in a real case to optimize an esterification reaction to achieve the maximum space-time-yield and the minimal E-factor. The algorithm identifies a clear trade-off between the two objectives.

多目标贝叶斯优化（MOBO）已被证明是一种很有前途的反应开发工具。然而，在实验和化学过程中，噪声通常是不可避免的，当噪声未知或显著时，寻找可靠的解决方案是具有挑战性的。在这项研究中，我们的重点是寻找一组最优的反应条件，使用多目标欧几里得期望分位数改进（MO-E-EQI）在噪声设置下。首先，将MO-E-EQI算法的性能与一些具有线性和对数线性异方差噪声结构和不同量级的计算机MOBO算法进行比较。注意到高噪声会降低MOBO算法的性能。MO-E-EQI在基于超容量的度量、覆盖度量和Pareto前端的解决方案数量方面表现出强大的性能。最后，应用MO-E-EQI对酯化反应进行优化，以达到最大的时空产率和最小的e因子。该算法明确了这两个目标之间的权衡。

{"title":"Multi-objective reaction optimization under uncertainties using expected quantile improvement","authors":"Jiyizhe Zhang , Daria Semochkina , Naoto Sugisawa , David C. Woods , Alexei A. Lapkin","doi":"10.1016/j.compchemeng.2024.108983","DOIUrl":"10.1016/j.compchemeng.2024.108983","url":null,"abstract":"<div><div>Multi-objective Bayesian optimization (MOBO) has shown to be a promising tool for reaction development. However, noise is usually inevitable in experimental and chemical processes, and finding reliable solutions is challenging when the noise is unknown or significant. In this study, we focus on finding a set of optimal reaction conditions using multi-objective Euclidian expected quantile improvement (MO-E-EQI) under noisy settings. First, the performance of MO-E-EQI is evaluated by comparing with some recent MOBO algorithms <em>in silico</em> with linear and log-linear heteroscedastic noise structures and different magnitudes. It is noticed that high noise can degrade the performance of MOBO algorithms. MO-E-EQI shows robust performance in terms of hypervolume-based metric, coverage metric and number of solutions on the Pareto front. Finally, MO-E-EQI is implemented in a real case to optimize an esterification reaction to achieve the maximum space-time-yield and the minimal E-factor. The algorithm identifies a clear trade-off between the two objectives.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"194 ","pages":"Article 108983"},"PeriodicalIF":3.9,"publicationDate":"2025-01-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143136740","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Optimal chemical reaction pathway for palm process residue recovery using Process Graph (P-graph) framework 利用过程图（P-graph）框架优化棕榈渣回收的化学反应途径

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

Computers & Chemical Engineering

Pub Date : 2025-01-05 DOI: 10.1016/j.compchemeng.2025.109000

Seen Ye Lim , Nishanth G. Chemmangattuvalappil , John Frederick D. Tapia , Ianatul Khoiroh , Pui Vun Chai , Lik Yin Ng

Oleochemical industry generates palm process residue during hydrogenation of fatty acids or methyl esters. This residue, comprising fatty alcohols and alkanes with overlapping boiling points, is challenging and costly to separate using conventional distillation. Efficient recovery of fatty alcohols for commercial use, while alkanes for jet fuel, lubricants, and gasoline are beneficial. A promising solution involves halogenating fatty alcohols into derivatives with distinct boiling points from alkanes, enabling efficient distillation. Thus, identifying chemical reaction pathways for fatty alcohols and halogenating agents that occurs spontaneously under optimal conditions is crucial for cost-effectiveness and sustainability. Utilizing P-graph framework with SSG + LP algorithm, 116 thermodynamically feasible pathways were generated and analyzed using Aspen Plus. The optimal pathway successfully separated C₁₂H₂₅OH from C₁₄H₃₀ and achieved a high conversion of 90.40% for C₁₂H₂₅Br. This pathway also produced valuable by-products such as C₄H₈BrOH and C₅H₁₁OH, generating higher revenue and demonstrating industrial feasibility.

油脂化学工业在脂肪酸或甲酯的氢化过程中产生棕榈过程残留物。这种残留物由脂肪醇和烷烃组成，沸点重叠，使用传统蒸馏分离是具有挑战性和昂贵的。高效回收脂肪醇用于商业用途，而烷烃用于喷气燃料、润滑油和汽油是有益的。一个有希望的解决方案是将脂肪醇卤化成与烷烃沸点不同的衍生物，从而实现高效蒸馏。因此，确定在最佳条件下自发发生的脂肪醇和卤化剂的化学反应途径对于成本效益和可持续性至关重要。利用P-graph框架和SSG + LP算法，生成了116条热力学可行路径，并使用Aspen Plus进行了分析。该优化途径成功地分离了C12H25OH和C14H30， C12H25Br的转化率高达90.40%。该途径还产生了C4H8BrOH和C5H11OH等有价值的副产物，产生了更高的收益，证明了工业可行性。

{"title":"Optimal chemical reaction pathway for palm process residue recovery using Process Graph (P-graph) framework","authors":"Seen Ye Lim , Nishanth G. Chemmangattuvalappil , John Frederick D. Tapia , Ianatul Khoiroh , Pui Vun Chai , Lik Yin Ng","doi":"10.1016/j.compchemeng.2025.109000","DOIUrl":"10.1016/j.compchemeng.2025.109000","url":null,"abstract":"<div><div>Oleochemical industry generates palm process residue during hydrogenation of fatty acids or methyl esters. This residue, comprising fatty alcohols and alkanes with overlapping boiling points, is challenging and costly to separate using conventional distillation. Efficient recovery of fatty alcohols for commercial use, while alkanes for jet fuel, lubricants, and gasoline are beneficial. A promising solution involves halogenating fatty alcohols into derivatives with distinct boiling points from alkanes, enabling efficient distillation. Thus, identifying chemical reaction pathways for fatty alcohols and halogenating agents that occurs spontaneously under optimal conditions is crucial for cost-effectiveness and sustainability. Utilizing P-graph framework with SSG + LP algorithm, 116 thermodynamically feasible pathways were generated and analyzed using Aspen Plus. The optimal pathway successfully separated C<sub>12</sub>H<sub>25</sub>OH from C<sub>14</sub>H<sub>30</sub> and achieved a high conversion of 90.40% for C<sub>12</sub>H<sub>25</sub>Br. This pathway also produced valuable by-products such as C<sub>4</sub>H<sub>8</sub>BrOH and C<sub>5</sub>H<sub>11</sub>OH, generating higher revenue and demonstrating industrial feasibility.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"194 ","pages":"Article 109000"},"PeriodicalIF":3.9,"publicationDate":"2025-01-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143136775","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

An ORP prediction model for acid wastewater sulfidation process based on improved extreme learning machine 基于改进极限学习机的酸性废水硫化过程ORP预测模型

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

Computers & Chemical Engineering

Pub Date : 2025-01-04 DOI: 10.1016/j.compchemeng.2025.108998

Hongqiu Zhu, Yixin Lv, Minghui Liu, Can Zhou

The flow of hydrogen sulfide is a crucial factor influencing the precipitation of heavy metals in acid wastewater. However, flow regulation in industrial environments often demonstrates lag. The oxidation–reduction potential (ORP) is closely linked to the flow of hydrogen sulfide. Consequently, this paper proposes an ORP prediction model that employs a double-layer improved particle swarm optimization (DLIPSO) and extreme learning machine (ELM). To overcome the limitation of particle swarm optimization (PSO) easily getting trapped in local optima, the oppositional-based learning (OBL) strategy and time-varying inertia weights are introduced to improve the search performance of the particles. Additionally, a double-layer particle swarm structure is utilized to identify the most effective combination of optimal structure and parameters for the ELM, maximizing its predictive performance. The proposed model is validated on a real dataset and compared with five other models. Experimental results indicate that the root mean square error (RMSE) of the proposed model decreased by 9.40 % to 49.76 % compared to the other models.

硫化氢的流动是影响酸性废水中重金属沉淀的关键因素。然而，工业环境中的流量调节往往表现出滞后性。氧化还原电位（ORP）与硫化氢的流动密切相关。因此，本文提出了一种采用双层改进粒子群优化（DLIPSO）和极限学习机（ELM）的ORP预测模型。为了克服粒子群算法容易陷入局部最优的缺点，引入了基于对立学习（OBL）策略和时变惯性权重来提高粒子群算法的搜索性能。此外，利用双层粒子群结构来识别最优结构和参数的最有效组合，最大限度地提高了ELM的预测性能。在实际数据集上对该模型进行了验证，并与其他五种模型进行了比较。实验结果表明，与其他模型相比，该模型的均方根误差（RMSE）降低了9.40% ~ 49.76%。

{"title":"An ORP prediction model for acid wastewater sulfidation process based on improved extreme learning machine","authors":"Hongqiu Zhu, Yixin Lv, Minghui Liu, Can Zhou","doi":"10.1016/j.compchemeng.2025.108998","DOIUrl":"10.1016/j.compchemeng.2025.108998","url":null,"abstract":"<div><div>The flow of hydrogen sulfide is a crucial factor influencing the precipitation of heavy metals in acid wastewater. However, flow regulation in industrial environments often demonstrates lag. The oxidation–reduction potential (ORP) is closely linked to the flow of hydrogen sulfide. Consequently, this paper proposes an ORP prediction model that employs a double-layer improved particle swarm optimization (DLIPSO) and extreme learning machine (ELM). To overcome the limitation of particle swarm optimization (PSO) easily getting trapped in local optima, the oppositional-based learning (OBL) strategy and time-varying inertia weights are introduced to improve the search performance of the particles. Additionally, a double-layer particle swarm structure is utilized to identify the most effective combination of optimal structure and parameters for the ELM, maximizing its predictive performance. The proposed model is validated on a real dataset and compared with five other models. Experimental results indicate that the root mean square error (RMSE) of the proposed model decreased by 9.40 % to 49.76 % compared to the other models.</div></div>","PeriodicalId":286,"journal":{"name":"Computers & Chemical Engineering","volume":"194 ","pages":"Article 108998"},"PeriodicalIF":3.9,"publicationDate":"2025-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143136595","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

Long term turnaround planning for an oil refinery using a MILP model 使用MILP模型的炼油厂长期周转计划

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

Computers & Chemical Engineering

Pub Date : 2025-01-04 DOI: 10.1016/j.compchemeng.2025.108999

Ricardo A.de O. Lima, Reginaldo Guirardello

This study presents a discrete-time mixed-integer linear programming (MILP) model to optimize long-term maintenance turnaround scheduling in an oil refinery focused on fuel production. Refineries are complex networks of integrated process units, and maintenance turnarounds, involving temporary shutdowns for inspection and repair, can significantly disrupt production and reduce revenues. The MILP model aims to minimize these disruptions by optimizing turnaround schedules while maintaining product supply and maximizing economic performance. The model incorporates flow, labor, resource, and planning constraints, allowing for different unit groupings and scenario simulations. Key outputs include the maintenance schedule, unit utilization rates, intermediate stock levels, production, manpower, and maintenance costs. The model serves as a decision-support tool for refining managers, enabling them to plan maintenance interventions that maximize operating profit while adhering to operational constraints.

本文提出了一个离散时间混合整数线性规划（MILP）模型来优化以燃料生产为重点的炼油厂的长期维修周转计划。炼油厂是一个复杂的集成过程单元网络，而维护周转（包括临时停工进行检查和维修）可能会严重破坏生产并减少收入。MILP模型旨在通过优化周转时间表，同时保持产品供应和最大化经济效益，最大限度地减少这些中断。该模型结合了流程、劳动力、资源和计划约束，允许不同的单元分组和场景模拟。关键输出包括维修计划、单位利用率、中间库存水平、生产、人力和维修成本。该模型可作为精炼管理人员的决策支持工具，使他们能够在遵守运营约束的情况下，计划维护干预措施，使运营利润最大化。

引用次数: 0