Computers & Industrial Engineering最新文献_第4页

Robust railway traffic routing problem under uncertain demand: an improved benders decomposition approach 不确定需求下的鲁棒铁路交通路径问题：一种改进的弯曲分解方法

IF 6.5 1区工程技术 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Industrial Engineering

Pub Date : 2026-01-28 DOI: 10.1016/j.cie.2026.111829

Yinan Zhao , Hanwen Jiang , Shiyi Gong

Robust traffic routing is an effective means of mitigating rail congestion and enhancing system resilience in emergencies and unexpected disruptions. However, there still lack a systematic theoretical system for uncertainty optimization in railway freight traffic organization, due to modeling complexity and limited availability of operational data. This study formulates a strategic robust railway freight traffic routing problem under demand uncertainty without requiring distributional assumptions. Demand fluctuations are captured through a symmetric uncertainty set with a robustness budget (price of robustness), yielding a computationally tractable deterministic reformulation for routing decisions. An improved Benders decomposition framework in which the master problem is approximately optimized by simulated annealing (SA). The numerical results on a series of expanding test networks demonstrate that the proposed SA-based Benders approach can serve as a highly qualified alternative solving approach when the computational resources are limited. Sensitivity analyses conducted on key algorithmic parameters indicate stable performance under moderate perturbations, supporting that the adopted baseline settings lie in a practically reasonable range.

稳健的交通路由是缓解铁路拥堵和增强系统在紧急情况和意外中断下的弹性的有效手段。然而，由于建模的复杂性和运营数据的可获得性有限，铁路货运组织的不确定性优化还缺乏系统的理论体系。本文研究了需求不确定条件下的战略鲁棒铁路货运路线问题。需求波动通过具有鲁棒性预算（鲁棒性价格）的对称不确定性集来捕获，从而为路由决策产生计算上可处理的确定性重新表述。采用模拟退火法对主问题进行近似优化，提出了一种改进的Benders分解框架。在一系列扩展的测试网络上的数值结果表明，在计算资源有限的情况下，基于sa的Benders方法可以作为一种高质量的替代求解方法。对关键算法参数进行的敏感性分析表明，在适度扰动下，算法性能稳定，支持所采用的基线设置处于实际合理的范围内。

{"title":"Robust railway traffic routing problem under uncertain demand: an improved benders decomposition approach","authors":"Yinan Zhao , Hanwen Jiang , Shiyi Gong","doi":"10.1016/j.cie.2026.111829","DOIUrl":"10.1016/j.cie.2026.111829","url":null,"abstract":"<div><div>Robust traffic routing is an effective means of mitigating rail congestion and enhancing system resilience in emergencies and unexpected disruptions. However, there still lack a systematic theoretical system for uncertainty optimization in railway freight traffic organization, due to modeling complexity and limited availability of operational data. This study formulates a strategic robust railway freight traffic routing problem under demand uncertainty without requiring distributional assumptions. Demand fluctuations are captured through a symmetric uncertainty set with a robustness budget (price of robustness), yielding a computationally tractable deterministic reformulation for routing decisions. An improved Benders decomposition framework in which the master problem is approximately optimized by simulated annealing (SA). The numerical results on a series of expanding test networks demonstrate that the proposed SA-based Benders approach can serve as a highly qualified alternative solving approach when the computational resources are limited. Sensitivity analyses conducted on key algorithmic parameters indicate stable performance under moderate perturbations, supporting that the adopted baseline settings lie in a practically reasonable range.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"214 ","pages":"Article 111829"},"PeriodicalIF":6.5,"publicationDate":"2026-01-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146190421","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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

Harmonizing sustainability and resiliency: A novel robust-stochastic decomposition approach for effective mask distribution and recycling 协调可持续性和弹性：一种新的鲁棒随机分解方法，用于有效的面罩分配和回收

IF 6.5 1区工程技术 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Industrial Engineering

Pub Date : 2026-01-28 DOI: 10.1016/j.cie.2026.111837

Mahsa Mohammadi , Babak Mohamadpour Tosarkani

This study aims to develop a novel graph theory-based clustering algorithm for large-scale logistics planning problems focusing on the distribution of face masks by various transportation modes under uncertainty. A robust multi-objective, mixed-integer linear programming model (MILPM) is utilized to handle imprecise parameters (e.g., demand and processing time). The proposed model supports decision-makers in designing a sustainable closed-loop supply chain network for the optimal face mask distribution under time window limitations. A sample average approximation methodology is applied to tackle the large-scale case study. Furthermore, a graph theory-based clustering algorithm is proposed to accelerate the scenario decomposition approach since it deals with less scenarios in comparison with sample average approximation and scenario decomposition. Silhouette analysis is conducted to measure the performance and accuracy of the generated clusters. Sensitivity analyses are implemented to validate the efficiency and applicability of the presented solution approach. A series of scenarios is set to represent supply chain network disruptions with unknown probabilities. The outcome of this study denotes the optimal flow of face masks and the optimum number of facilities at the time of the COVID-19 outbreak in Toronto, Canada.

针对不确定条件下不同运输方式口罩分配的大规模物流规划问题，提出一种新的基于图论的聚类算法。利用鲁棒多目标混合整数线性规划模型（MILPM）处理不精确参数（如需求和加工时间）。该模型支持决策者在时间窗口限制下设计可持续的闭环供应链网络，以实现口罩的最优分配。采用样本平均近似方法处理大规模案例研究。此外，提出了一种基于图论的聚类算法，与样本平均近似和场景分解相比，它处理的场景更少，从而加快了场景分解的速度。通过剪影分析来衡量生成的聚类的性能和准确性。通过灵敏度分析验证了所提出的求解方法的有效性和适用性。一系列场景被设置为代表未知概率的供应链网络中断。这项研究的结果表明，在加拿大多伦多发生COVID-19疫情时，口罩的最佳流量和设施的最佳数量。

{"title":"Harmonizing sustainability and resiliency: A novel robust-stochastic decomposition approach for effective mask distribution and recycling","authors":"Mahsa Mohammadi , Babak Mohamadpour Tosarkani","doi":"10.1016/j.cie.2026.111837","DOIUrl":"10.1016/j.cie.2026.111837","url":null,"abstract":"<div><div>This study aims to develop a novel graph theory-based clustering algorithm for large-scale logistics planning problems focusing on the distribution of face masks by various transportation modes under uncertainty. A robust multi-objective, mixed-integer linear programming model <em>(MILPM)</em> is utilized to handle imprecise parameters (e.g., demand and processing time). The proposed model supports decision-makers in designing a sustainable closed-loop supply chain network for the optimal face mask distribution under time window limitations. A sample average approximation methodology is applied to tackle the large-scale case study. Furthermore, a graph theory-based clustering algorithm is proposed to accelerate the scenario decomposition approach since it deals with less scenarios in comparison with sample average approximation and scenario decomposition. Silhouette analysis is conducted to measure the performance and accuracy of the generated clusters. Sensitivity analyses are implemented to validate the efficiency and applicability of the presented solution approach. A series of scenarios is set to represent supply chain network disruptions with unknown probabilities. The outcome of this study denotes the optimal flow of face masks and the optimum number of facilities at the time of the COVID-19 outbreak in Toronto, Canada.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"214 ","pages":"Article 111837"},"PeriodicalIF":6.5,"publicationDate":"2026-01-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146081359","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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

Optimal control policy for combined production–maintenance and multivariate quality monitoring of an imperfect manufacturing system with replenishment and assignable causes 具有补货和可分配原因的不完善制造系统的生产维护和多变量质量监控的最优控制策略

IF 6.5 1区工程技术 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Industrial Engineering

Pub Date : 2026-01-28 DOI: 10.1016/j.cie.2026.111871

Qiang Wan , Yanni An , Mei Zhu

Current research that integrates the three functions of production planning, maintenance, and quality monitoring suffers from three significant shortcomings: (1) Assume that there is only one identifiable source of variation within the process. Due to the inherent complexity characteristic of most real-world manufacturing operations, the simplified assumption of a single assignable cause is seldom observed in actual industrial settings. When the actual assignable cause of a process shift does not match the one anticipated in quality monitoring, performance may be subpar in both economic and statistical terms. (2) Monitoring a single quality characteristic, however, in actual conditions, multiple process quality characteristics should be monitored simultaneously. (3) To simplify the model, the buffer inventory time is established at the start of the cycle. However, if the buffer is loaded too early, it may result in excess inventory holding expenses. To handle these shortcomings, this work establishes an integration scheme for production, multivariate statistical process monitoring and maintenance planning that considers dynamical replenishment and multiple assignable causes. Under economic–statistical quality constraints, a customized genetic algorithm is employed to optimize the expected total cost of each process cycle. In the comparative study, the proposed model is compared with integrated models using a single assignable cause, MEWMA, and MCUSUM charts, highlighting its superior economic and statistical performance. Finally, a design of experiments (DOE)-based sensitivity analysis is carried out on the principal process parameters and the average total cost per cycle.

目前将生产计划、维护和质量监控三个功能集成在一起的研究存在三个显著缺陷：(1)假设过程中只有一个可识别的变异源。由于大多数现实世界制造操作的固有复杂性，在实际工业环境中很少观察到单一可分配原因的简化假设。当过程转移的实际可分配原因与质量监控中预期的原因不匹配时，从经济和统计角度来看，绩效可能低于标准。(2)监测单一质量特征，但在实际条件下，应同时监测多个过程质量特征。(3)为了简化模型，在周期开始时建立缓冲库存时间。但是，如果缓冲区加载得太早，可能会导致库存持有费用过多。为了解决这些问题，本文建立了一个生产、多元统计过程监控和维修计划的集成方案，该方案考虑了动态补货和多种可分配原因。在经济统计质量约束下，采用自定义遗传算法对各工序周期的期望总成本进行优化。在比较研究中，将所提出的模型与使用单一可分配原因、MEWMA和mccusum图表的综合模型进行了比较，突出了其优越的经济和统计性能。最后，对主要工艺参数和每周期平均总成本进行了基于试验设计（DOE）的灵敏度分析。

{"title":"Optimal control policy for combined production–maintenance and multivariate quality monitoring of an imperfect manufacturing system with replenishment and assignable causes","authors":"Qiang Wan , Yanni An , Mei Zhu","doi":"10.1016/j.cie.2026.111871","DOIUrl":"10.1016/j.cie.2026.111871","url":null,"abstract":"<div><div>Current research that integrates the three functions of production planning, maintenance, and quality monitoring suffers from three significant shortcomings: (1) Assume that there is only one identifiable source of variation within the process. Due to the inherent complexity characteristic of most real-world manufacturing operations, the simplified assumption of a single assignable cause is seldom observed in actual industrial settings. When the actual assignable cause of a process shift does not match the one anticipated in quality monitoring, performance may be subpar in both economic and statistical terms. (2) Monitoring a single quality characteristic, however, in actual conditions, multiple process quality characteristics should be monitored simultaneously. (3) To simplify the model, the buffer inventory time is established at the start of the cycle. However, if the buffer is loaded too early, it may result in excess inventory holding expenses. To handle these shortcomings, this work establishes an integration scheme for production, multivariate statistical process monitoring and maintenance planning that considers dynamical replenishment and multiple assignable causes. Under economic–statistical quality constraints, a customized genetic algorithm is employed to optimize the expected total cost of each process cycle. In the comparative study, the proposed model is compared with integrated models using a single assignable cause, MEWMA, and MCUSUM charts, highlighting its superior economic and statistical performance. Finally, a design of experiments (DOE)-based sensitivity analysis is carried out on the principal process parameters and the average total cost per cycle.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"214 ","pages":"Article 111871"},"PeriodicalIF":6.5,"publicationDate":"2026-01-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146081360","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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 multi-stage stochastic model for sustainable semiconductor manufacturing 可持续半导体制造的多阶段随机模型

IF 6.5 1区工程技术 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Industrial Engineering

Pub Date : 2026-01-27 DOI: 10.1016/j.cie.2026.111869

Anshuman Kumar, S.P. Sarmah

Semiconductor manufacturing systems are subject to significant operational uncertainties stemming from fluctuating customer demand, variable supplier lead times and machine-level disruptions. This study presents a multi-stage stochastic programming framework that integrates production planning, procurement scheduling, inventory control and emissions management within a unified decision-making model. The framework explicitly incorporates environmental regulations through periodic emission thresholds and tool cleaning constraints while accounting for sourcing risks through supplier classification and diversification. By modelling uncertainty through a scenario-based approach, the proposed method enables both anticipatory and adaptive decisions that enhance system robustness. The model is implemented using mixed-integer programming techniques and validated through computational experiments based on empirically motivated scenarios. Results demonstrate improved cost efficiency, service level adherence, and regulatory compliance compared to deterministic baselines. Sensitivity analysis highlights key trade-offs, showing that stricter emission caps can increase total costs, while supplier diversification helps mitigate disruption risks. The results underscore the value of stochastic programming in capturing the complex interdependencies in semiconductor supply chains and provide a rigorous decision-support tool for managing uncertainty in high-precision manufacturing systems.

由于客户需求的波动、供应商交货时间的变化和机器层面的中断，半导体制造系统受到重大运营不确定性的影响。本文提出了一个多阶段随机规划框架，将生产计划、采购调度、库存控制和排放管理集成在一个统一的决策模型中。该框架通过定期排放阈值和工具清洁限制明确纳入环境法规，同时通过供应商分类和多样化考虑采购风险。通过基于场景的方法对不确定性进行建模，所提出的方法可以实现预期和自适应决策，从而增强系统的鲁棒性。该模型使用混合整数规划技术实现，并通过基于经验动机场景的计算实验进行验证。结果表明，与确定基线相比，成本效率、服务水平依从性和法规遵从性得到了提高。敏感性分析强调了关键的权衡，表明更严格的排放上限可以增加总成本，而供应商多样化有助于减轻中断风险。研究结果强调了随机规划在捕获半导体供应链中复杂的相互依赖性方面的价值，并为管理高精度制造系统中的不确定性提供了严格的决策支持工具。

{"title":"A multi-stage stochastic model for sustainable semiconductor manufacturing","authors":"Anshuman Kumar, S.P. Sarmah","doi":"10.1016/j.cie.2026.111869","DOIUrl":"10.1016/j.cie.2026.111869","url":null,"abstract":"<div><div>Semiconductor manufacturing systems are subject to significant operational uncertainties stemming from fluctuating customer demand, variable supplier lead times and machine-level disruptions. This study presents a multi-stage stochastic programming framework that integrates production planning, procurement scheduling, inventory control and emissions management within a unified decision-making model. The framework explicitly incorporates environmental regulations through periodic emission thresholds and tool cleaning constraints while accounting for sourcing risks through supplier classification and diversification. By modelling uncertainty through a scenario-based approach, the proposed method enables both anticipatory and adaptive decisions that enhance system robustness. The model is implemented using mixed-integer programming techniques and validated through computational experiments based on empirically motivated scenarios. Results demonstrate improved cost efficiency, service level adherence, and regulatory compliance compared to deterministic baselines. Sensitivity analysis highlights key trade-offs, showing that stricter emission caps can increase total costs, while supplier diversification helps mitigate disruption risks. The results underscore the value of stochastic programming in capturing the complex interdependencies in semiconductor supply chains and provide a rigorous decision-support tool for managing uncertainty in high-precision manufacturing systems.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"214 ","pages":"Article 111869"},"PeriodicalIF":6.5,"publicationDate":"2026-01-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146190357","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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

Reinforcement learning for train timetable rescheduling under perturbation: A general value-based approach 扰动下列车时刻表重新调度的强化学习：一种通用的基于值的方法

IF 6.5 1区工程技术 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Industrial Engineering

Pub Date : 2026-01-27 DOI: 10.1016/j.cie.2026.111867

Pu Zhang , Lingyun Meng , Yongqiu Zhu , Jianrui Miao , Xiaojie Luan , Zhengwen Liao

This paper proposes a value-based deep reinforcement learning approach that is capable of handling train timetable rescheduling under both disturbed and disrupted situations. A railway environment is constructed to simulate the problem as a Markov decision process, where the optimization objective is integrated into the reward module and various constraints are incorporated into the conflict detection and avoidance module. To address the challenges of sparse rewards and large action space with limited legal actions, a value-based algorithm framework is proposed to efficiently select and effectively evaluate actions. Through the designed simulation and training procedures, the proposed approach is tested on several disturbance and disruption cases based on a real-world instance (i.e. a Chinese high-speed railway corridor). Experimental results show that the proposed method can obtain high-quality solutions within a reasonable computing time, and also outperforms handcrafted rules in terms of the optimality of solutions. Furthermore, the proposed method exhibits promising generalization capabilities in homogeneous perturbation scenarios (disturbance scenarios and disruption scenarios that share either the same affected location and start time or the same affected location and disrupted duration).

本文提出了一种基于值的深度强化学习方法，该方法能够处理在干扰和中断情况下的列车时刻表重新调度。构建铁路环境，将问题模拟为马尔可夫决策过程，将优化目标集成到奖励模块中，将各种约束集成到冲突检测与规避模块中。为了解决奖励稀疏、行动空间大、合法行动有限的问题，提出了一种基于值的算法框架来高效地选择和评估行动。通过设计的仿真和训练程序，在基于现实实例（即中国高速铁路走廊）的几种干扰和中断案例中对所提出的方法进行了测试。实验结果表明，该方法可以在合理的计算时间内获得高质量的解，并且在解的最优性方面优于手工规则。此外，所提出的方法在同质扰动场景（干扰场景和中断场景共享相同的受影响位置和开始时间或相同的受影响位置和中断持续时间）中显示出有希望的泛化能力。

{"title":"Reinforcement learning for train timetable rescheduling under perturbation: A general value-based approach","authors":"Pu Zhang , Lingyun Meng , Yongqiu Zhu , Jianrui Miao , Xiaojie Luan , Zhengwen Liao","doi":"10.1016/j.cie.2026.111867","DOIUrl":"10.1016/j.cie.2026.111867","url":null,"abstract":"<div><div>This paper proposes a value-based deep reinforcement learning approach that is capable of handling train timetable rescheduling under both disturbed and disrupted situations. A railway environment is constructed to simulate the problem as a Markov decision process, where the optimization objective is integrated into the reward module and various constraints are incorporated into the conflict detection and avoidance module. To address the challenges of sparse rewards and large action space with limited legal actions, a value-based algorithm framework is proposed to efficiently select and effectively evaluate actions. Through the designed simulation and training procedures, the proposed approach is tested on several disturbance and disruption cases based on a real-world instance (i.e. a Chinese high-speed railway corridor). Experimental results show that the proposed method can obtain high-quality solutions within a reasonable computing time, and also outperforms handcrafted rules in terms of the optimality of solutions. Furthermore, the proposed method exhibits promising generalization capabilities in homogeneous perturbation scenarios (disturbance scenarios and disruption scenarios that share either the same affected location and start time or the same affected location and disrupted duration).</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"214 ","pages":"Article 111867"},"PeriodicalIF":6.5,"publicationDate":"2026-01-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146190360","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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

Joint optimization of vending machine deployment and shelf display design with synchronized merchandise replenishment 联合优化自动售货机部署和货架展示设计，同步商品补充

IF 6.5 1区工程技术 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Industrial Engineering

Pub Date : 2026-01-25 DOI: 10.1016/j.cie.2026.111848

Kung-Jeng Wang , Natalia Febri

Vending machines (VMs) serve as an important aspect of automated retail, delivering both flexibility in operations and convenience for consumers. However, as VM networks expand, managers face growing logistical challenges in determining optimal deployment locations, product selection and allocation, and restocking schedules. This study proposes a novel bi-layer optimization framework that jointly optimizes deployment, product selection and allocation, and a synchronized replenishment cycle. To address the complexity of this large-scale combinatorial problem, we develop a hybrid Tabu Search and Evolution Strategy (TS-ES) algorithm. Extensive experiments show that the synchronized replenishment cycle yields better performance than the independent cycle. Comparative analysis demonstrates that the hybrid TS-ES algorithm consistently achieves higher objective values than standalone TS, genetic algorithm (GA), random search (RS), and iterative local search (ILS) across various problem scales. This research contributes to the current body of knowledge by introducing a comprehensive framework that improves VM operational performance and serves as a practical resource for optimizing the logistics and profitability within VM networks.

自动售货机（vm）是自动化零售的一个重要方面，为消费者提供操作的灵活性和便利性。然而，随着虚拟机网络的扩展，管理人员在确定最佳部署位置、产品选择和分配以及重新进货计划方面面临着越来越多的后勤挑战。本研究提出了一种新的双层优化框架，共同优化部署、产品选择和分配以及同步补货周期。为了解决这种大规模组合问题的复杂性，我们开发了一种混合禁忌搜索和进化策略（TS-ES）算法。大量实验表明，同步补货周期比独立补货周期具有更好的性能。对比分析表明，在不同的问题尺度上，混合TS- es算法均比独立TS、遗传算法（GA）、随机搜索（RS）和迭代局部搜索（ILS）获得更高的目标值。本研究通过引入一个全面的框架来提高虚拟机的运行性能，并作为虚拟机网络中优化物流和盈利能力的实用资源，为当前的知识体系做出了贡献。

{"title":"Joint optimization of vending machine deployment and shelf display design with synchronized merchandise replenishment","authors":"Kung-Jeng Wang , Natalia Febri","doi":"10.1016/j.cie.2026.111848","DOIUrl":"10.1016/j.cie.2026.111848","url":null,"abstract":"<div><div>Vending machines (VMs) serve as an important aspect of automated retail, delivering both flexibility in operations and convenience for consumers. However, as VM networks expand, managers face growing logistical challenges in determining optimal deployment locations, product selection and allocation, and restocking schedules. This study proposes a novel bi-layer optimization framework that jointly optimizes deployment, product selection and allocation, and a synchronized replenishment cycle. To address the complexity of this large-scale combinatorial problem, we develop a hybrid Tabu Search and Evolution Strategy (TS-ES) algorithm. Extensive experiments show that the synchronized replenishment cycle yields better performance than the independent cycle. Comparative analysis demonstrates that the hybrid TS-ES algorithm consistently achieves higher objective values than standalone TS, genetic algorithm (GA), random search (RS), and iterative local search (ILS) across various problem scales. This research contributes to the current body of knowledge by introducing a comprehensive framework that improves VM operational performance and serves as a practical resource for optimizing the logistics and profitability within VM networks.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"214 ","pages":"Article 111848"},"PeriodicalIF":6.5,"publicationDate":"2026-01-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146081469","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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 scenario-adaptive optimization model for circular intertwined supply network design under uncertainty 不确定条件下循环交织供电网络设计的场景自适应优化模型

IF 6.5 1区工程技术 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Industrial Engineering

Pub Date : 2026-01-25 DOI: 10.1016/j.cie.2026.111849

Mohaddeseh Roshan , Jessica Olivares-Aguila , Waguih ElMaraghy

Intertwined supply networks are collaborative, cross-industry supply chains characterized by a high level of interconnectedness among their entities. This study demonstrates that integrating circular economy principles into such networks can cut greenhouse gas emissions by 4.97% and reduce total system costs by 11.03%, while strengthening economic efficiency and social responsibility under uncertainty. To realize these improvements, a novel multi-objective non-linear mixed-integer mathematical model is proposed with an embedded scenario differentiation mechanism that enables evaluation of configurations, from traditional decentralized supply chains to complex intertwined networks with varying levels of circularity, within a unified analytical model. The objective functions are to minimize system costs and greenhouse gas emissions and maximize social responsibility for optimal location decisions under uncertainty. The proposed model is first verified using the AUGMECON-2 method and validated via a case study of an intertwined pharmaceutical-bioplastic supply network, complemented by numerical experiments, sensitivity analyses, and a comparative study using the Grey Wolf Optimizer for large-scale instances.

交织的供应网络是协作的、跨行业的供应链，其特点是其实体之间的高度互联性。研究表明，将循环经济原则融入此类网络可减少4.97%的温室气体排放，降低11.03%的系统总成本，同时增强不确定性下的经济效率和社会责任。为了实现这些改进，提出了一种新的多目标非线性混合整数数学模型，该模型具有嵌入式场景区分机制，可以在统一的分析模型中对配置进行评估，从传统的分散供应链到具有不同循环水平的复杂交织网络。不确定条件下最优选址决策的目标函数是系统成本和温室气体排放最小化，社会责任最大化。首先使用AUGMECON-2方法验证了所提出的模型，并通过一个相互交织的制药-生物塑料供应网络的案例研究进行了验证，并辅以数值实验、敏感性分析和使用灰狼优化器进行大规模实例的比较研究。

{"title":"A scenario-adaptive optimization model for circular intertwined supply network design under uncertainty","authors":"Mohaddeseh Roshan , Jessica Olivares-Aguila , Waguih ElMaraghy","doi":"10.1016/j.cie.2026.111849","DOIUrl":"10.1016/j.cie.2026.111849","url":null,"abstract":"<div><div>Intertwined supply networks are collaborative, cross-industry supply chains characterized by a high level of interconnectedness among their entities. This study demonstrates that integrating circular economy principles into such networks can cut greenhouse gas emissions by 4.97% and reduce total system costs by 11.03%, while strengthening economic efficiency and social responsibility under uncertainty. To realize these improvements, a novel multi-objective non-linear mixed-integer mathematical model is proposed with an embedded scenario differentiation mechanism that enables evaluation of configurations, from traditional decentralized supply chains to complex intertwined networks with varying levels of circularity, within a unified analytical model. The objective functions are to minimize system costs and greenhouse gas emissions and maximize social responsibility for optimal location decisions under uncertainty. The proposed model is first verified using the AUGMECON-2 method and validated via a case study of an intertwined pharmaceutical-bioplastic supply network, complemented by numerical experiments, sensitivity analyses, and a comparative study using the Grey Wolf Optimizer for large-scale instances.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"214 ","pages":"Article 111849"},"PeriodicalIF":6.5,"publicationDate":"2026-01-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146081472","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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

Simulation-based optimization of industrial symbiosis under carbon regulations: towards sustainable and resilient production networks 碳监管下基于模拟的工业共生优化：走向可持续和有弹性的生产网络

IF 6.5 1区工程技术 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Industrial Engineering

Pub Date : 2026-01-25 DOI: 10.1016/j.cie.2026.111847

Lina Aboueljinane , Maroua Sbiti , Youness Frichi

While Industrial Symbiosis (IS) is a cornerstone of the circular economy, its operational viability remains vulnerable to market volatility and regulatory shifts. This study investigates the resilience and sustainability of a symbiotic system under five carbon regulation mechanisms. A capacitated joint production planning problem is formulated as a Mixed-Integer Linear Program (MILP) and compared against a Simulation-Based Optimization framework using Differential Evolution (DE). To evaluate global robustness, we conducted a massive stress-test across 5000 scenarios generated via Latin Hypercube Sampling (LHS) and analyzed using Explainable AI (Random Forest). The results reveal a critical “Structural Fragility”: the deterministic MILP exhibits a 60% failure rate, categorized either as Mathematical Infeasibility (solver incapacity) or Physical Operational Failure (inventory overflows during simulation), driven by capacity bottlenecks and demand surges. In contrast, the proposed framework guarantees 99.76% % feasibility by dynamically adjusting safety stocks, identifying robust “best-effort” solutions. A multi-objective Pareto analysis further quantifies the trade-offs, revealing a Shadow Price of Resilience and an Environmental Rebound Effect. Among regulatory mechanisms, Cap-and-Trade emerges as the “smartest“ policy, enabling a dynamic arbitrage capability, where the system intelligently switches between production and carbon trading based on market signals. This study contributes a unified, data-driven framework for designing resilient, low-carbon manufacturing systems capable of withstanding real-world uncertainty.

虽然工业共生（IS）是循环经济的基石，但其运营可行性仍然容易受到市场波动和监管变化的影响。本研究探讨了五种碳调控机制下共生系统的恢复力和可持续性。将有能力联合生产计划问题表述为混合整数线性规划（MILP），并与基于仿真的差分进化优化框架进行了比较。为了评估全局稳健性，我们通过拉丁超立方体采样（LHS）对5000个场景进行了大规模的压力测试，并使用可解释人工智能（随机森林）进行了分析。结果揭示了一个关键的“结构脆弱性”：确定性的MILP显示出60%的失败率，被分类为数学上的不可行性（求解器无能）或物理操作失败（模拟期间的库存溢出），由产能瓶颈和需求激增驱动。相比之下，提议的框架通过动态调整安全库存，确定稳健的“尽力而为”解决方案，保证99.76%的可行性。多目标帕累托分析进一步量化了权衡，揭示了弹性的影子价格和环境反弹效应。在监管机制中，总量管制与交易成为“最聪明”的政策，实现了动态套利能力，系统根据市场信号在生产和碳交易之间智能切换。该研究为设计能够承受现实世界不确定性的弹性低碳制造系统提供了统一的数据驱动框架。

{"title":"Simulation-based optimization of industrial symbiosis under carbon regulations: towards sustainable and resilient production networks","authors":"Lina Aboueljinane , Maroua Sbiti , Youness Frichi","doi":"10.1016/j.cie.2026.111847","DOIUrl":"10.1016/j.cie.2026.111847","url":null,"abstract":"<div><div>While Industrial Symbiosis (IS) is a cornerstone of the circular economy, its operational viability remains vulnerable to market volatility and regulatory shifts. This study investigates the resilience and sustainability of a symbiotic system under five carbon regulation mechanisms. A capacitated joint production planning problem is formulated as a Mixed-Integer Linear Program (MILP) and compared against a Simulation-Based Optimization framework using Differential Evolution (DE). To evaluate global robustness, we conducted a massive stress-test across 5000 scenarios generated via Latin Hypercube Sampling (LHS) and analyzed using Explainable AI (Random Forest). The results reveal a critical “Structural Fragility”: the deterministic MILP exhibits a 60% failure rate, categorized either as Mathematical Infeasibility (solver incapacity) or Physical Operational Failure (inventory overflows during simulation), driven by capacity bottlenecks and demand surges. In contrast, the proposed framework guarantees 99.76% % feasibility by dynamically adjusting safety stocks, identifying robust “best-effort” solutions. A multi-objective Pareto analysis further quantifies the trade-offs, revealing a Shadow Price of Resilience and an Environmental Rebound Effect. Among regulatory mechanisms, Cap-and-Trade emerges as the “smartest“ policy, enabling a dynamic arbitrage capability, where the system intelligently switches between production and carbon trading based on market signals. This study contributes a unified, data-driven framework for designing resilient, low-carbon manufacturing systems capable of withstanding real-world uncertainty.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"214 ","pages":"Article 111847"},"PeriodicalIF":6.5,"publicationDate":"2026-01-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146081475","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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

PG-DGPRL: A physics-guided and uncertainty-aware feedforward force-admittance control method for soft-bodied operational robots PG-DGPRL：一种物理引导和不确定性感知的软体操作机器人前馈力导纳控制方法

IF 6.5 1区工程技术 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Industrial Engineering

Pub Date : 2026-01-25 DOI: 10.1016/j.cie.2026.111838

Yurong Guo, Jie Zhang, Junliang Wang

To address the challenges of control latency and uncertainty in force control tasks involving soft-bodied operational robots, this paper proposes a physics-guided and uncertainty-aware feedforward force-admittance control method, referred to as PG-DGPRL. First, a feedforward admittance control strategy with online estimation of force-tracking error is introduced. By forecasting future deviations in interaction forces, the controller parameters are adjusted in advance, thereby mitigating the adverse effects of control latency. Second, a physics-informed deep Gaussian process reinforcement learning (DGPRL) method is developed. The policy is represented as a Gaussian distribution, and the constitutive mechanical model of flexible materials is embedded as a physical constraint, enabling both uncertainty awareness and physical consistency verification. In addition, a composite loss function is designed to achieve joint optimization among policy generation by the actor network, uncertainty estimation by the DGP, and physical regularization by a physics-informed neural network (PINN). Finally, experiments are conducted under various environmental damping and interaction force conditions, comparing PG-DGPRL with baseline methods including CAC, Ada-CAC, and A2C. The results indicate that, under the current experimental conditions, PG-DGPRL achieves superior trajectory tracking and force control performance, exhibiting strong stability and generalization capability.

针对软体操作机器人在力控制任务中存在的控制延迟和不确定性问题，提出了一种物理引导、不确定性感知的前馈力导纳控制方法PG-DGPRL。首先，介绍了一种在线估计力跟踪误差的前馈导纳控制策略。通过预测未来交互力的偏差，提前调整控制器参数，从而减轻控制延迟的不利影响。其次，提出了一种基于物理的深度高斯过程强化学习（DGPRL）方法。该策略以高斯分布表示，柔性材料的本构力学模型作为物理约束嵌入，实现了不确定性意识和物理一致性验证。此外，设计了一个复合损失函数，以实现行动者网络的策略生成、DGP的不确定性估计和物理信息神经网络（PINN）的物理正则化之间的联合优化。最后，在各种环境阻尼和相互作用力条件下进行实验，将PG-DGPRL与CAC、Ada-CAC和A2C等基准方法进行比较。结果表明，在目前的实验条件下，PG-DGPRL具有较好的轨迹跟踪和力控制性能，具有较强的稳定性和泛化能力。

{"title":"PG-DGPRL: A physics-guided and uncertainty-aware feedforward force-admittance control method for soft-bodied operational robots","authors":"Yurong Guo, Jie Zhang, Junliang Wang","doi":"10.1016/j.cie.2026.111838","DOIUrl":"10.1016/j.cie.2026.111838","url":null,"abstract":"<div><div>To address the challenges of control latency and uncertainty in force control tasks involving soft-bodied operational robots, this paper proposes a physics-guided and uncertainty-aware feedforward force-admittance control method, referred to as PG-DGPRL. First, a feedforward admittance control strategy with online estimation of force-tracking error is introduced. By forecasting future deviations in interaction forces, the controller parameters are adjusted in advance, thereby mitigating the adverse effects of control latency. Second, a physics-informed deep Gaussian process reinforcement learning (DGPRL) method is developed. The policy is represented as a Gaussian distribution, and the constitutive mechanical model of flexible materials is embedded as a physical constraint, enabling both uncertainty awareness and physical consistency verification. In addition, a composite loss function is designed to achieve joint optimization among policy generation by the actor network, uncertainty estimation by the DGP, and physical regularization by a physics-informed neural network (PINN). Finally, experiments are conducted under various environmental damping and interaction force conditions, comparing PG-DGPRL with baseline methods including CAC, Ada-CAC, and A2C. The results indicate that, under the current experimental conditions, PG-DGPRL achieves superior trajectory tracking and force control performance, exhibiting strong stability and generalization capability.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"214 ","pages":"Article 111838"},"PeriodicalIF":6.5,"publicationDate":"2026-01-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146190419","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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

High-performance concrete compressive strength prediction using Soft computing 基于软计算的高性能混凝土抗压强度预测

IF 6.5 1区工程技术 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Industrial Engineering

Pub Date : 2026-01-24 DOI: 10.1016/j.cie.2026.111840

Yushuai Zhang , Xiao Li , Yunqiang Wang , Zhenyu Wang , Chuan Shi , Jiyong Lei , Licheng Sun

Accurate prediction of the compressive strength of high-performance concrete is of great significance for concrete engineering design and quality control. To address the limitations of traditional prediction methods characterized by insufficient accuracy and inadequate interpretability, this study proposes an intelligent prediction methodology that integrates domain knowledge-driven feature engineering with Light Gradient Boosting Machine (LightGBM) optimized by Tree-structured Parzen Estimator (TPE). Based on the high-performance concrete dataset, this study utilized concrete materials science theory to construct a composite feature system comprising five major categories encompassing 12 engineering features: water-binder ratio, cementitious materials, aggregate proportion, admixture effects, and comprehensive performance. Through an incremental feature selection strategy, the system was optimized to yield 10 critical features. The TPE Bayesian optimization algorithm was employed to conduct hyperparameter tuning for the LightGBM model, and SHapley Additive exPlanations (SHAP) methodology was utilized to perform interpretability analysis. The experimental results demonstrate that the optimized model achieved superior performance on the test set with Root Mean Square Error (RMSE) of 3.371 MPa, coefficient of determination (R²) of 0.961, Mean Absolute Error (MAE) of 2.446 MPa, and Mean Absolute Percentage Error (MAPE) of 9.6 %. The proposed method outperformed advanced approaches such as eXtreme Gradient Boosting and Random Forest, comprehensively validating the effectiveness of the proposed methodology. SHAP analysis revealed that Age and Effective_Water_Cement_Ratio were the most critical predictive factors, with the model-learned patterns demonstrating high consistency with concrete hydration theory. This research provides a precise and interpretable prediction tool for designing high-performance concrete with significant practical engineering value.

高性能混凝土抗压强度的准确预测对混凝土工程设计和质量控制具有重要意义。针对传统预测方法精度不足、可解释性不足的局限性，提出了一种将领域知识驱动特征工程与基于树结构Parzen Estimator （TPE）优化的Light Gradient Boosting Machine （LightGBM）相结合的智能预测方法。本研究以高性能混凝土数据集为基础，运用混凝土材料学理论，构建了包含水胶比、胶凝材料、骨料比、外加剂效应、综合性能等5大类12项工程特征的复合特征体系。通过增量特征选择策略，对系统进行优化，得到10个关键特征。采用TPE贝叶斯优化算法对LightGBM模型进行超参数调优，采用SHapley加性解释（SHAP）方法进行可解释性分析。实验结果表明，优化后的模型在测试集上取得了较好的性能，均方根误差（RMSE）为3.371 MPa，决定系数（R2）为0.961，平均绝对误差（MAE）为2.446 MPa，平均绝对百分比误差（MAPE）为9.6%。该方法优于极端梯度增强和随机森林等先进方法，全面验证了该方法的有效性。SHAP分析显示，龄期和有效水水泥比是最关键的预测因素，模型学习模式与混凝土水化理论高度一致。本研究为高性能混凝土的设计提供了一种精确、可解释的预测工具，具有重要的工程实用价值。

{"title":"High-performance concrete compressive strength prediction using Soft computing","authors":"Yushuai Zhang , Xiao Li , Yunqiang Wang , Zhenyu Wang , Chuan Shi , Jiyong Lei , Licheng Sun","doi":"10.1016/j.cie.2026.111840","DOIUrl":"10.1016/j.cie.2026.111840","url":null,"abstract":"<div><div>Accurate prediction of the compressive strength of high-performance concrete is of great significance for concrete engineering design and quality control. To address the limitations of traditional prediction methods characterized by insufficient accuracy and inadequate interpretability, this study proposes an intelligent prediction methodology that integrates domain knowledge-driven feature engineering with Light Gradient Boosting Machine (LightGBM) optimized by Tree-structured Parzen Estimator (TPE). Based on the high-performance concrete dataset, this study utilized concrete materials science theory to construct a composite feature system comprising five major categories encompassing 12 engineering features: water-binder ratio, cementitious materials, aggregate proportion, admixture effects, and comprehensive performance. Through an incremental feature selection strategy, the system was optimized to yield 10 critical features. The TPE Bayesian optimization algorithm was employed to conduct hyperparameter tuning for the LightGBM model, and SHapley Additive exPlanations (SHAP) methodology was utilized to perform interpretability analysis. The experimental results demonstrate that the optimized model achieved superior performance on the test set with Root Mean Square Error (RMSE) of 3.371 MPa, coefficient of determination (R<sup>2</sup>) of 0.961, Mean Absolute Error (MAE) of 2.446 MPa, and Mean Absolute Percentage Error (MAPE) of 9.6 %. The proposed method outperformed advanced approaches such as eXtreme Gradient Boosting and Random Forest, comprehensively validating the effectiveness of the proposed methodology. SHAP analysis revealed that Age and Effective_Water_Cement_Ratio were the most critical predictive factors, with the model-learned patterns demonstrating high consistency with concrete hydration theory. This research provides a precise and interpretable prediction tool for designing high-performance concrete with significant practical engineering value.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"214 ","pages":"Article 111840"},"PeriodicalIF":6.5,"publicationDate":"2026-01-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146190359","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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