Computers & Industrial Engineering最新文献_第10页

A deep reinforcement learning approach for integrated optimization of train scheduling and rolling stock circulation planning 列车调度与车辆循环规划综合优化的深度强化学习方法

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

Computers & Industrial Engineering

Pub Date : 2025-12-25 DOI: 10.1016/j.cie.2025.111784

Xiaoli Zhao, Dewei Li, Xinyu Bao

This study proposes a deep reinforcement learning-based optimization framework for integrated train scheduling and rolling stock circulation planning under dynamic passenger demand. The problem is formulated as a Markov decision process (MDP) with a hybrid action space that simultaneously captures continuous timetable decisions and discrete rolling stock allocations. The objective is to minimize passenger waiting time and operator costs while adhering to complex operational constraints. To address the challenge of simultaneously coordinating continuous and discrete decision variables in a high-dimensional operational context, we adopt a Hybrid Proximal Policy Optimization (HPPO) algorithm, incorporating separate actor networks for discrete and continuous actions, and employing constraint-handling techniques such as action masking and action space embedding. Furthermore, a potential-based reward shaping function is introduced to enhance learning efficiency by addressing issues of sparse and delayed rewards. The proposed approach is validated on the Beijing Metro Changping Line. Experimental results demonstrate that the HPPO algorithm effectively improves system efficiency and policy robustness.

本文提出了一种基于深度强化学习的动态客运需求下列车调度与车辆循环综合规划优化框架。该问题被表述为具有混合动作空间的马尔可夫决策过程（MDP），同时捕获连续的时间表决策和离散的机车车辆分配。目标是在遵守复杂的操作约束的同时，最大限度地减少乘客等待时间和运营商成本。为了解决在高维操作环境中同时协调连续和离散决策变量的挑战，我们采用了混合近端策略优化（HPPO）算法，为离散和连续动作结合单独的行动者网络，并采用约束处理技术，如动作掩蔽和动作空间嵌入。此外，引入基于电位的奖励塑造函数，通过解决奖励的稀疏和延迟问题来提高学习效率。该方法在北京地铁昌平线上得到了验证。实验结果表明，HPPO算法有效地提高了系统效率和策略鲁棒性。

{"title":"A deep reinforcement learning approach for integrated optimization of train scheduling and rolling stock circulation planning","authors":"Xiaoli Zhao, Dewei Li, Xinyu Bao","doi":"10.1016/j.cie.2025.111784","DOIUrl":"10.1016/j.cie.2025.111784","url":null,"abstract":"<div><div>This study proposes a deep reinforcement learning-based optimization framework for integrated train scheduling and rolling stock circulation planning under dynamic passenger demand. The problem is formulated as a Markov decision process (MDP) with a hybrid action space that simultaneously captures continuous timetable decisions and discrete rolling stock allocations. The objective is to minimize passenger waiting time and operator costs while adhering to complex operational constraints. To address the challenge of simultaneously coordinating continuous and discrete decision variables in a high-dimensional operational context, we adopt a Hybrid Proximal Policy Optimization (HPPO) algorithm, incorporating separate actor networks for discrete and continuous actions, and employing constraint-handling techniques such as action masking and action space embedding. Furthermore, a potential-based reward shaping function is introduced to enhance learning efficiency by addressing issues of sparse and delayed rewards. The proposed approach is validated on the Beijing Metro Changping Line. Experimental results demonstrate that the HPPO algorithm effectively improves system efficiency and policy robustness.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"213 ","pages":"Article 111784"},"PeriodicalIF":6.5,"publicationDate":"2025-12-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145885432","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

AI-enabled green production–inventory with dual channels, warranty returns, and blockchain carbon trading 人工智能支持的绿色生产——双渠道库存、保修退货和区块链碳交易

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

Computers & Industrial Engineering

Pub Date : 2025-12-24 DOI: 10.1016/j.cie.2025.111785

Prabal Das , Nabendu Sen , Ali Akbar Shaikh

Growing environmental regulations and rising consumer awareness have made it crucial for manufacturers to design inventory systems that strike a balance between profitability and sustainability. This work develops a dynamic green production–inventory model with price- and warranty-sensitive dual-channel demand, preservation investment, product deterioration, warranty-driven remanufacturing, and carbon-emission constraints. The model is evaluated across three cases: (i) a baseline system with warranty-based returns; (ii) a blockchain-enabled cap-and-trade mechanism; and (iii) an AI-guided adaptive preservation strategy that responds to real-time demand and emission levels. The system is formulated using nonlinear differential equations and solved via the Artificial Ecosystem-Based Optimizer (AEO). Case 1 yields an average profit of INR 2997.31. relative to Case 1, Case 2 increases average profit by

\approx

180.72% and reduces emissions by

\approx

9.52% through carbon-credit trading. Case 3 achieves an average profit of INR 8416.59-＋180.80% vs. Case 1-while reducing emissions by

\approx

21.43%. Under matched computational budgets, mainstream metaheuristics (PSO/GA/DE) reach a similar neighborhood of solutions, while AEO exhibits stable convergence with minimal tuning, corroborating robustness. Sensitivity analysis highlights demand elasticity and preservation-investment parameters as dominant profit drivers, and carbon pricing as a key environmental lever. The framework offers a scalable and adaptable decision-support tool for integrating AI, blockchain, and green investment into circular supply-chain design.

越来越多的环保法规和日益提高的消费者意识，使得制造商设计出能够在盈利能力和可持续性之间取得平衡的库存系统变得至关重要。本文建立了一个动态绿色生产库存模型，该模型包含价格和保修敏感的双渠道需求、保藏投资、产品劣化、保修驱动的再制造和碳排放约束。该模型在三种情况下进行评估：(i)具有基于保修的回报的基线系统；（ii）支持区块链的限额与交易机制；（iii）人工智能引导的适应性保护策略，以响应实时需求和排放水平。该系统采用非线性微分方程，并通过基于人工生态系统的优化器（AEO）进行求解。案例1的平均利润为2997.31印度卢比。相对于案例1，案例2通过碳信用交易，平均利润增加约180.72%，排放量减少约9.52%。与案例1相比，案例3实现了8416.59-+180.80%的平均利润，同时减少了约21.43%的排放量。在计算预算匹配的情况下，主流元启发式算法（PSO/GA/DE）得到了相似的邻域解，而AEO算法在最小调优下表现出稳定的收敛性，证实了鲁棒性。敏感性分析强调需求弹性和保护投资参数是主要的利润驱动因素，而碳定价是关键的环境杠杆。该框架提供了一个可扩展和适应性强的决策支持工具，用于将人工智能、bbb和绿色投资整合到循环供应链设计中。

{"title":"AI-enabled green production–inventory with dual channels, warranty returns, and blockchain carbon trading","authors":"Prabal Das , Nabendu Sen , Ali Akbar Shaikh","doi":"10.1016/j.cie.2025.111785","DOIUrl":"10.1016/j.cie.2025.111785","url":null,"abstract":"<div><div>Growing environmental regulations and rising consumer awareness have made it crucial for manufacturers to design inventory systems that strike a balance between profitability and sustainability. This work develops a dynamic green production–inventory model with price- and warranty-sensitive dual-channel demand, preservation investment, product deterioration, warranty-driven remanufacturing, and carbon-emission constraints. The model is evaluated across three cases: (i) a baseline system with warranty-based returns; (ii) a blockchain-enabled cap-and-trade mechanism; and (iii) an AI-guided adaptive preservation strategy that responds to real-time demand and emission levels. The system is formulated using nonlinear differential equations and solved via the Artificial Ecosystem-Based Optimizer (AEO). Case 1 yields an average profit of INR 2997.31. relative to Case 1, Case 2 increases <em>average</em> profit by <span><math><mo>≈</mo></math></span>180.72% and reduces emissions by <span><math><mo>≈</mo></math></span>9.52% through carbon-credit trading. Case 3 achieves an average profit of INR 8416.59-<strong>＋180.80% vs. Case 1</strong>-while reducing emissions by <span><math><mo>≈</mo></math></span>21.43%. Under matched computational budgets, mainstream metaheuristics (PSO/GA/DE) reach a similar neighborhood of solutions, while AEO exhibits stable convergence with minimal tuning, corroborating robustness. Sensitivity analysis highlights demand elasticity and preservation-investment parameters as dominant profit drivers, and carbon pricing as a key environmental lever. The framework offers a scalable and adaptable decision-support tool for integrating AI, blockchain, and green investment into circular supply-chain design.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"213 ","pages":"Article 111785"},"PeriodicalIF":6.5,"publicationDate":"2025-12-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145884900","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

Comparative estimation techniques for exponential-Rayleigh models under adaptive type-II progressive censoring 自适应ii型渐进式滤波下指数-瑞利模型的比较估计技术

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

Computers & Industrial Engineering

Pub Date : 2025-12-24 DOI: 10.1016/j.cie.2025.111788

M.S. Kotb , M.Z. Raqab

This paper presents a comprehensive analysis of adaptive type-II progressive censored data under the two-parameter exponential-Rayleigh distribution. Maximum likelihood and Bayesian methods are applied to estimate the model parameters. We construct asymptotic confidence intervals, equal-tailed Bayesian credible intervals, and highest posterior density (HPD) intervals. The methodology is illustrated through the analysis of a real data set and further assessed via Monte Carlo simulations to evaluate the performance of the proposed estimation procedures.

本文综合分析了双参数指数瑞利分布下的自适应ii型渐进式截尾数据。采用极大似然和贝叶斯方法对模型参数进行估计。我们构造了渐近置信区间、等尾贝叶斯可信区间和最高后验密度（HPD）区间。通过对真实数据集的分析说明了该方法，并通过蒙特卡罗模拟进一步评估了所提出的估计程序的性能。

引用次数: 0

Data-driven order batching policy: Focusing on the fulfillment center of Kurly 数据驱动的订单批处理策略：关注Kurly的履行中心

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

Computers & Industrial Engineering

Pub Date : 2025-12-24 DOI: 10.1016/j.cie.2025.111767

Juyoung Wang, Soohyung Park, Byungju Goo, Dongyun Kang, Soojee Kim, Jaehyeong Choi, Choonoh Lee

Order batching is a key operation in e-commerce fulfillment centers, exerting major influence on labor costs, throughput, and overall facility performance. In high-volume grocery warehouses, such as those operated by Kurly Inc., efficient batching is particularly crucial to meet narrow delivery windows and keep a high-level customer satisfaction. Traditional approaches often aim to minimize travel distance for item picking tasks or makespan, but Kurly’s primary bottleneck arises from the accumulation of totes (plastic bins) on conveyors, delaying downstream packing and limiting daily processing capacity. To address this challenge, we propose a data-driven order batching framework that directly targets tote number reduction by embedding a linear regression model into an integer program. The regression predicts how many totes a candidate batch will require, allowing us to optimize order-to-batch assignments with minimal tote usage. Tests with real operational data show that our hybrid strategy combining a genetic algorithm-based order batching policy and Gurobi’s no relaxation heuristic reduces tote congestion by up to 92% compared to Kurly’s legacy randomized batching algorithm. These advances expedite packing, increase order throughput, and improve resource utilization. Overall, our findings illustrate how data-driven objectives and off-the-shelf solvers can be successfully leveraged in large-scale industrial contexts to address pervasive bottlenecks in modern e-commerce warehousing.

订单批处理是电子商务履行中心的一项关键操作，对劳动力成本、吞吐量和整体设施性能产生重大影响。在库利公司（Kurly Inc.）运营的大容量杂货仓库中，高效的批处理对于满足狭窄的交货窗口和保持高水平的客户满意度尤为重要。传统方法的目标通常是尽量减少拣货任务的运输距离或最大完工时间，但Kurly的主要瓶颈来自输送带上的手提袋（塑料箱）的堆积，延迟了下游包装，限制了日常处理能力。为了应对这一挑战，我们提出了一个数据驱动的订单批处理框架，该框架通过将线性回归模型嵌入到整数程序中，直接针对减少订单数量。回归预测候选批将需要多少个手提袋，允许我们以最小的手提袋使用量优化订单到批的分配。使用实际操作数据进行的测试表明，与Kurly的传统随机批处理算法相比，我们的混合策略结合了基于遗传算法的顺序批处理策略和Gurobi的无松弛启发式算法，可减少高达92%的包拥塞。这些进步加快了包装，增加了订单吞吐量，提高了资源利用率。总的来说，我们的研究结果说明了数据驱动的目标和现成的解决方案如何在大规模工业环境中成功地利用，以解决现代电子商务仓储中普遍存在的瓶颈。

{"title":"Data-driven order batching policy: Focusing on the fulfillment center of Kurly","authors":"Juyoung Wang, Soohyung Park, Byungju Goo, Dongyun Kang, Soojee Kim, Jaehyeong Choi, Choonoh Lee","doi":"10.1016/j.cie.2025.111767","DOIUrl":"10.1016/j.cie.2025.111767","url":null,"abstract":"<div><div>Order batching is a key operation in e-commerce fulfillment centers, exerting major influence on labor costs, throughput, and overall facility performance. In high-volume grocery warehouses, such as those operated by Kurly Inc., efficient batching is particularly crucial to meet narrow delivery windows and keep a high-level customer satisfaction. Traditional approaches often aim to minimize travel distance for item picking tasks or makespan, but Kurly’s primary bottleneck arises from the accumulation of totes (plastic bins) on conveyors, delaying downstream packing and limiting daily processing capacity. To address this challenge, we propose a data-driven order batching framework that directly targets tote number reduction by embedding a linear regression model into an integer program. The regression predicts how many totes a candidate batch will require, allowing us to optimize order-to-batch assignments with minimal tote usage. Tests with real operational data show that our hybrid strategy combining a genetic algorithm-based order batching policy and Gurobi’s no relaxation heuristic reduces tote congestion by up to 92% compared to Kurly’s legacy randomized batching algorithm. These advances expedite packing, increase order throughput, and improve resource utilization. Overall, our findings illustrate how data-driven objectives and off-the-shelf solvers can be successfully leveraged in large-scale industrial contexts to address pervasive bottlenecks in modern e-commerce warehousing.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"213 ","pages":"Article 111767"},"PeriodicalIF":6.5,"publicationDate":"2025-12-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145841687","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

Multi-objective optimization of a truck–drone delivery system for fair and efficient humanitarian logistics under disruption and disinformation 卡车-无人机运输系统在干扰和虚假信息下的公平高效人道主义物流多目标优化

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

Computers & Industrial Engineering

Pub Date : 2025-12-24 DOI: 10.1016/j.cie.2025.111786

Ramin Talebi Khameneh, Nafiseh Ghorbani-Renani, Jose Emmanuel Ramirez-Marquez

Humanitarian logistics systems face significant challenges due to infrastructure damage, accessibility constraints, and the need for equitable and timely aid delivery. This paper introduces the Multi-Truck and UAV Routing Problem (MTURP), a novel bi-objective optimization model that coordinates trucks and drones to support last-mile delivery under disrupted conditions. The two objectives are to minimize the delivery time gap to improve fairness and reduce total travel distance to enhance efficiency. Unlike existing truck–drone routing models that optimize latency or cost alone, our formulation introduces a fairness objective based on a minimax criterion, minimizing the temporal gap between the earliest and latest deliveries to ensure equitable service. The model also incorporates socially informed prioritization using community vulnerability data (SVI) as equity weights, linking operational optimization with social resilience considerations. Given the problem’s NP-hard nature, we develop a solution approach based on the Non-dominated Sorting Genetic Algorithm II (NSGA-II) and benchmark its performance against an

ϵ

-constraint exact method using Gurobi. Results demonstrate that NSGA-II can achieve near-optimal solutions within 3%–10% of the Pareto front while significantly reducing computational time, especially for larger instances. Two case studies validate the framework: (i) an urban flood in Hoboken, NJ, and (ii) a rural disaster in Hopkins County, KY, where SVI-based weights are used solely to prioritize aid equitably across communities under uncertain information conditions. This approach emphasizes fairness without excluding any demand nodes or inferring disinformation, aligning with humanitarian principles of inclusive and needs-based allocation. Overall, the proposed framework highlights the practical value of integrating collaborative truck–drone operations with multi-objective optimization for scalable, fair, and disruption-resilient humanitarian logistics.

由于基础设施受损、可及性受限，以及公平和及时提供援助的需求，人道主义物流系统面临重大挑战。本文介绍了多卡车和无人机路径问题（MTURP），这是一种新的双目标优化模型，用于协调卡车和无人机在中断条件下支持最后一英里的交付。两个目标是最小化交货时间差距以提高公平性，减少总行程距离以提高效率。与现有的仅优化延迟或成本的卡车-无人机路线模型不同，我们的公式引入了基于最小和最大标准的公平目标，最大限度地减少最早和最晚交付之间的时间差距，以确保公平的服务。该模型还结合了使用社区脆弱性数据（SVI）作为公平权重的社会知情优先级，将运营优化与社会弹性考虑联系起来。考虑到问题的NP-hard性质，我们开发了一种基于非支配排序遗传算法II （NSGA-II）的解决方法，并使用Gurobi对其性能与ϵ-constraint精确方法进行了基准测试。结果表明，NSGA-II可以在Pareto前沿的3%-10%的范围内获得接近最优的解决方案，同时显著减少了计算时间，特别是对于较大的实例。两个案例研究验证了该框架：(i)新泽西州霍博肯市的城市洪水，（ii）肯塔基州霍普金斯县的农村灾害，其中基于svi的权重仅用于在不确定信息条件下公平地优先考虑社区之间的援助。这种方法强调公平，不排除任何需求节点或推断虚假信息，符合包容性和基于需求的分配的人道主义原则。总体而言，所提出的框架强调了将卡车-无人机协同操作与多目标优化相结合的实用价值，以实现可扩展、公平和抗破坏的人道主义物流。

{"title":"Multi-objective optimization of a truck–drone delivery system for fair and efficient humanitarian logistics under disruption and disinformation","authors":"Ramin Talebi Khameneh, Nafiseh Ghorbani-Renani, Jose Emmanuel Ramirez-Marquez","doi":"10.1016/j.cie.2025.111786","DOIUrl":"10.1016/j.cie.2025.111786","url":null,"abstract":"<div><div>Humanitarian logistics systems face significant challenges due to infrastructure damage, accessibility constraints, and the need for equitable and timely aid delivery. This paper introduces the Multi-Truck and UAV Routing Problem (MTURP), a novel bi-objective optimization model that coordinates trucks and drones to support last-mile delivery under disrupted conditions. The two objectives are to minimize the delivery time gap to improve fairness and reduce total travel distance to enhance efficiency. Unlike existing truck–drone routing models that optimize latency or cost alone, our formulation introduces a fairness objective based on a minimax criterion, minimizing the temporal gap between the earliest and latest deliveries to ensure equitable service. The model also incorporates socially informed prioritization using community vulnerability data (SVI) as equity weights, linking operational optimization with social resilience considerations. Given the problem’s NP-hard nature, we develop a solution approach based on the Non-dominated Sorting Genetic Algorithm II (NSGA-II) and benchmark its performance against an <span><math><mi>ϵ</mi></math></span>-constraint exact method using Gurobi. Results demonstrate that NSGA-II can achieve near-optimal solutions within 3%–10% of the Pareto front while significantly reducing computational time, especially for larger instances. Two case studies validate the framework: (i) an urban flood in Hoboken, NJ, and (ii) a rural disaster in Hopkins County, KY, where SVI-based weights are used solely to prioritize aid equitably across communities under uncertain information conditions. This approach emphasizes fairness without excluding any demand nodes or inferring disinformation, aligning with humanitarian principles of inclusive and needs-based allocation. Overall, the proposed framework highlights the practical value of integrating collaborative truck–drone operations with multi-objective optimization for scalable, fair, and disruption-resilient humanitarian logistics.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"213 ","pages":"Article 111786"},"PeriodicalIF":6.5,"publicationDate":"2025-12-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145885438","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

Accounting for Within-Part Variability in Measurement System Analysis 测量系统分析中部分内变异性的核算

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

Computers & Industrial Engineering

Pub Date : 2025-12-23 DOI: 10.1016/j.cie.2025.111766

Mahmut Onur Karaman , Murat Caner Testik , Murat Kulahci , Stefan Steiner , Yusuf Emre Arslan

Measurement system analysis is often performed in industry to check the quality of a measurement system. When assessing the capability of measurement systems, two key components of measurement error are commonly considered: repeatability and reproducibility. Experimental designs are generally used to quantify these error components, and the resulting investigation is referred to as a Gauge Repeatability and Reproducibility (Gauge R&R) study. Yet another possible source of error that is often overlooked is within-part variability, which can be a significant source of variability in the measurements of parts. Investigation of this error source requires modifications to both the experimental setup and the effects model for quantifying the components of variability. In this study, we analyze the effects of various measurement assessment plans on the Gauge R&R metric when within-part variation is present. Alternative measurement assessment plans and their ANOVA models are proposed. As common geometric forms in manufacturing processes, cylindrical and rectangular prism shaped parts are considered. Using computer simulations of parts having within-part variability, it is shown that neglecting within-part variation can cause misleading conclusions from an MSA study, and as a consequence, practitioners may not recognize the actual performance of their measurement system. Our conclusions are justified by a real world case study for a cylindrical part.

测量系统分析在工业中经常用于检查测量系统的质量。在评估测量系统的能力时，通常要考虑测量误差的两个关键组成部分：可重复性和再现性。实验设计通常用于量化这些误差成分，由此产生的调查被称为量规重复性和再现性（Gauge R&；R）研究。然而，另一个经常被忽视的可能的误差来源是部件内部的可变性，这可能是部件测量中可变性的一个重要来源。对这种误差源的研究需要修改实验设置和用于量化变异成分的影响模型。在本研究中，我们分析了当存在部分内变化时，各种测量评估计划对量规R&；R度量的影响。提出了备选测量评估方案及其方差分析模型。圆柱形和矩形棱柱形零件是制造过程中常见的几何形状。使用具有部分内变异性的部件的计算机模拟，表明忽略部分内变异性可能导致从MSA研究中得出误导性结论，因此，从业者可能无法识别其测量系统的实际性能。我们的结论是由一个圆柱形零件的实际案例研究证明。

{"title":"Accounting for Within-Part Variability in Measurement System Analysis","authors":"Mahmut Onur Karaman , Murat Caner Testik , Murat Kulahci , Stefan Steiner , Yusuf Emre Arslan","doi":"10.1016/j.cie.2025.111766","DOIUrl":"10.1016/j.cie.2025.111766","url":null,"abstract":"<div><div>Measurement system analysis is often performed in industry to check the quality of a measurement system. When assessing the capability of measurement systems, two key components of measurement error are commonly considered: repeatability and reproducibility. Experimental designs are generally used to quantify these error components, and the resulting investigation is referred to as a Gauge Repeatability and Reproducibility (Gauge R&R) study. Yet another possible source of error that is often overlooked is within-part variability, which can be a significant source of variability in the measurements of parts. Investigation of this error source requires modifications to both the experimental setup and the effects model for quantifying the components of variability. In this study, we analyze the effects of various measurement assessment plans on the Gauge R&R metric when within-part variation is present. Alternative measurement assessment plans and their ANOVA models are proposed. As common geometric forms in manufacturing processes, cylindrical and rectangular prism shaped parts are considered. Using computer simulations of parts having within-part variability, it is shown that neglecting within-part variation can cause misleading conclusions from an MSA study, and as a consequence, practitioners may not recognize the actual performance of their measurement system. Our conclusions are justified by a real world case study for a cylindrical part.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"213 ","pages":"Article 111766"},"PeriodicalIF":6.5,"publicationDate":"2025-12-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145841686","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

Online process monitoring under quality data scarcity: Self-starting truncated EWMA schemes for time between events 高质量数据稀缺下的在线过程监控：事件间隔时间的自启动截断EWMA方案

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

Computers & Industrial Engineering

Pub Date : 2025-12-22 DOI: 10.1016/j.cie.2025.111777

FuPeng Xie , JingWei Liu , JiaCai Huang , YuXing Dai , Quan Sun , XueLong Hu , Philippe Castagliola

This study proposes one-sided self-starting truncated EWMA (SST-EWMA) control charts for effective high-quality process monitoring in situations where extensive in-control time-between-events (TBE) observations are unavailable. By constructing a pivot quantity and establishing variable mappings, a self-starting framework specifically tailored for Gamma distributed TBE observations is developed. The integration of a variable truncation mechanism into this framework further enhances sensitivity to small to moderate process shifts. To investigate the detection properties of the proposed schemes, simulation were conducted to examine the effects of the shape parameter

α

, the number of reference TBE observations

M

, and the smoothing parameter

λ

on the average time to signal (

ATS

). Based on the simulation results, guidelines are provided for achieving

ATS

performance comparable to that of the corresponding known-parameter schemes. Comparative analysis demonstrates that, although slightly inferior to the one-sided TEWMA TBE charts under known parameters, the proposed charts exhibit superior adaptability in scenarios with scarce TBE data, and also outperform the existing self-starting EWMA TBE chart, validating the effectiveness of the variable truncation mechanism. Finally, two case studies are presented to illustrate the practical implementation of the proposed control charts in industrial engineering applications.

本研究提出了单侧自启动截断EWMA （SST-EWMA）控制图，用于在无法获得大量控制时间间隔（TBE）观测的情况下进行有效的高质量过程监控。通过构造一个枢轴量和建立变量映射，开发了一个专门为Gamma分布TBE观测量身定制的自启动框架。将可变截断机制集成到该框架中，进一步提高了对小到中等过程移位的敏感性。为了研究所提方案的检测特性，通过仿真研究了形状参数α、参考TBE观测数M和平滑参数λ对平均到信号时间（ATS）的影响。根据仿真结果，给出了实现ATS性能与相应的已知参数方案相当的准则。对比分析表明，虽然在已知参数下，本文提出的图略逊于单侧的EWMA TBE图，但在TBE数据稀缺的场景下，该图表现出更强的适应性，也优于现有的自启动EWMA TBE图，验证了变量截断机制的有效性。最后，提出了两个案例来说明所提出的控制图在工业工程应用中的实际实现。

{"title":"Online process monitoring under quality data scarcity: Self-starting truncated EWMA schemes for time between events","authors":"FuPeng Xie , JingWei Liu , JiaCai Huang , YuXing Dai , Quan Sun , XueLong Hu , Philippe Castagliola","doi":"10.1016/j.cie.2025.111777","DOIUrl":"10.1016/j.cie.2025.111777","url":null,"abstract":"<div><div>This study proposes one-sided self-starting truncated EWMA (SST-EWMA) control charts for effective high-quality process monitoring in situations where extensive in-control time-between-events (TBE) observations are unavailable. By constructing a pivot quantity and establishing variable mappings, a self-starting framework specifically tailored for Gamma distributed TBE observations is developed. The integration of a variable truncation mechanism into this framework further enhances sensitivity to small to moderate process shifts. To investigate the detection properties of the proposed schemes, simulation were conducted to examine the effects of the shape parameter <span><math><mi>α</mi></math></span>, the number of reference TBE observations <span><math><mi>M</mi></math></span>, and the smoothing parameter <span><math><mi>λ</mi></math></span> on the average time to signal (<span><math><mi>ATS</mi></math></span>). Based on the simulation results, guidelines are provided for achieving <span><math><mi>ATS</mi></math></span> performance comparable to that of the corresponding known-parameter schemes. Comparative analysis demonstrates that, although slightly inferior to the one-sided TEWMA TBE charts under known parameters, the proposed charts exhibit superior adaptability in scenarios with scarce TBE data, and also outperform the existing self-starting EWMA TBE chart, validating the effectiveness of the variable truncation mechanism. Finally, two case studies are presented to illustrate the practical implementation of the proposed control charts in industrial engineering applications.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"213 ","pages":"Article 111777"},"PeriodicalIF":6.5,"publicationDate":"2025-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145841685","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 novel process monitoring method for multivariate autocorrelated mixed-type data based on transformer model integrated with the feature-enhancement learning method 一种基于变压器模型与特征增强学习方法相结合的多变量自相关混合型数据过程监测方法

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

Computers & Industrial Engineering

Pub Date : 2025-12-22 DOI: 10.1016/j.cie.2025.111776

Li Xue , Sen Feng , Tianci Zhao , Zhen He , Yuanzhong Jia , Tianye He

In the context of big data-driven intelligent manufacturing, process data often exhibit high dimensionality, multiple variables, and complex correlations. When these data comprise multiple variable types with unknown distributions, traditional parametric control charts struggle to address these challenges effectively. This study leverages the transformer model to monitor mixed continuous, count, and categorical data types in multivariate autocorrelated processes, proposing a feature-enhancement learning monitoring method. The experimental results demonstrate that the transformer model integrated with the proposed feature-enhancement learning method outperforms traditional monitoring approaches, such as residual control charts and

T^{2}

control charts, as well as other models, such as back propagation (BP), convolutional neural network (CNN), recurrent neural network (RNN), gated recurrent unit (GRU) and long short-term memory (LSTM), and also two traditional control charts constructed based on statistical methods for monitoring mixed-type data. The method’s effectiveness is further validated through a case study in semiconductor manufacturing. This study provides a theoretical foundation for applying deep learning technology to monitor multivariate autocorrelated processes.

在大数据驱动的智能制造背景下，过程数据往往呈现高维、多变量、复杂关联的特征。当这些数据包含具有未知分布的多种变量类型时，传统的参数控制图难以有效地解决这些挑战。本研究利用变压器模型监测多元自相关过程中的混合连续、计数和分类数据类型，提出了一种特征增强学习监测方法。实验结果表明，结合特征增强学习方法的变压器模型优于残差控制图和T2控制图等传统监测方法，也优于反向传播（BP）、卷积神经网络（CNN）、递归神经网络（RNN）、门控递归单元（GRU）和长短期记忆（LSTM）等其他模型。基于统计方法构造了两种传统的控制图，用于监测混合类型数据。通过半导体制造的实例研究，进一步验证了该方法的有效性。本研究为应用深度学习技术监测多变量自相关过程提供了理论基础。

{"title":"A novel process monitoring method for multivariate autocorrelated mixed-type data based on transformer model integrated with the feature-enhancement learning method","authors":"Li Xue , Sen Feng , Tianci Zhao , Zhen He , Yuanzhong Jia , Tianye He","doi":"10.1016/j.cie.2025.111776","DOIUrl":"10.1016/j.cie.2025.111776","url":null,"abstract":"<div><div>In the context of big data-driven intelligent manufacturing, process data often exhibit high dimensionality, multiple variables, and complex correlations. When these data comprise multiple variable types with unknown distributions, traditional parametric control charts struggle to address these challenges effectively. This study leverages the transformer model to monitor mixed continuous, count, and categorical data types in multivariate autocorrelated processes, proposing a feature-enhancement learning monitoring method. The experimental results demonstrate that the transformer model integrated with the proposed feature-enhancement learning method outperforms traditional monitoring approaches, such as residual control charts and <span><math><msup><mrow><mi>T</mi></mrow><mrow><mn>2</mn></mrow></msup></math></span> control charts, as well as other models, such as back propagation (BP), convolutional neural network (CNN), recurrent neural network (RNN), gated recurrent unit (GRU) and long short-term memory (LSTM), and also two traditional control charts constructed based on statistical methods for monitoring mixed-type data. The method’s effectiveness is further validated through a case study in semiconductor manufacturing. This study provides a theoretical foundation for applying deep learning technology to monitor multivariate autocorrelated processes.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"213 ","pages":"Article 111776"},"PeriodicalIF":6.5,"publicationDate":"2025-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145885431","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

Multi-objective scheduling for complex assembly shops considering multiple human factors 考虑多人为因素的复杂装配车间多目标调度

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

Computers & Industrial Engineering

Pub Date : 2025-12-20 DOI: 10.1016/j.cie.2025.111773

Huiting Li , Jiapeng Zhang , Xiaodi Chen , Haoxin Guo , Jianhua Liu , Cunbo Zhuang

The advancement of Industry 5.0 has driven a growing body of research that examines the impact of human factors on production processes. However, studies that simultaneously consider multiple types of human factors remain scarce. In this study, a comprehensive set of human factors, including workers’ skill proficiency, fatigue levels, interpersonal dynamics, and work experience, is incorporated into the assembly scheduling framework. Based on these considerations, the multi-objective scheduling problem in complex product assembly shops with parallel teams is investigated, with optimization objectives including makespan, transportation time, total waiting time, and team workload imbalance. To address this problem, an improved non-dominated sorting genetic algorithm is proposed. The algorithm features enhancement strategies, such as a destruction-reconstruction approach for optimizing the initial population and an improved evolutionary process. The proposed algorithm is evaluated against alternative algorithms using four case studies derived from actual production scenarios. The results demonstrate that the proposed method achieves superior solution quality and efficiency.

工业5.0的进步推动了越来越多的研究，研究人为因素对生产过程的影响。然而，同时考虑多种人为因素的研究仍然很少。本研究将工人的技能熟练程度、疲劳程度、人际关系动态和工作经验等人为因素纳入装配调度框架。在此基础上，研究了具有并行团队的复杂产品装配车间的多目标调度问题，优化目标包括完工时间、运输时间、总等待时间和团队工作量不平衡。为了解决这一问题，提出了一种改进的非支配排序遗传算法。该算法具有增强策略，如用于优化初始种群的破坏-重建方法和改进的进化过程。该算法通过从实际生产场景中导出的四个案例研究，对备选算法进行了评估。结果表明，该方法具有较好的求解质量和求解效率。

{"title":"Multi-objective scheduling for complex assembly shops considering multiple human factors","authors":"Huiting Li , Jiapeng Zhang , Xiaodi Chen , Haoxin Guo , Jianhua Liu , Cunbo Zhuang","doi":"10.1016/j.cie.2025.111773","DOIUrl":"10.1016/j.cie.2025.111773","url":null,"abstract":"<div><div>The advancement of Industry 5.0 has driven a growing body of research that examines the impact of human factors on production processes. However, studies that simultaneously consider multiple types of human factors remain scarce. In this study, a comprehensive set of human factors, including workers’ skill proficiency, fatigue levels, interpersonal dynamics, and work experience, is incorporated into the assembly scheduling framework. Based on these considerations, the multi-objective scheduling problem in complex product assembly shops with parallel teams is investigated, with optimization objectives including makespan, transportation time, total waiting time, and team workload imbalance. To address this problem, an improved non-dominated sorting genetic algorithm is proposed. The algorithm features enhancement strategies, such as a destruction-reconstruction approach for optimizing the initial population and an improved evolutionary process. The proposed algorithm is evaluated against alternative algorithms using four case studies derived from actual production scenarios. The results demonstrate that the proposed method achieves superior solution quality and efficiency.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"213 ","pages":"Article 111773"},"PeriodicalIF":6.5,"publicationDate":"2025-12-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145841694","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

Demand prediction for bike-sharing systems: A spatial and semantic modeling approach for enhanced accuracy and operational efficiency 自行车共享系统的需求预测：提高准确性和运行效率的空间和语义建模方法

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

Computers & Industrial Engineering

Pub Date : 2025-12-20 DOI: 10.1016/j.cie.2025.111775

Juntao Wu , Jiahui Feng , Jie Fang , Hefu Liu

The exponential growth of Bike-Sharing Systems (BSS) has introduced complex challenges in supply–demand management, where imbalances frequently lead to resource wastage and reduced user satisfaction. While Graph Neural Networks (GNNs) have become a mainstream tool for demand forecasting, existing methodologies predominantly rely on static geographic proximity, failing to capture the latent semantic dependencies driven by actual riding behaviors. To bridge this gap, this paper proposes a novel Spatial-Semantic Graph Attention Neural Network (SSGAN). Unlike traditional models, SSGAN constructs a semantic adjacency matrix using DTW to quantify the shape similarity between station inflow and outflow patterns, thereby capturing non-Euclidean correlations beyond physical distance. Furthermore, a Gated Multi-Head Attention mechanism is designed to dynamically weigh these semantic relationships by integrating external covariates (e.g., weather), allowing the model to adapt to time-varying contexts. Crucially, to align prediction accuracy with decision effectiveness, the model employs a dual-stream architecture that fuses inflow and outflow features to better reflect net inventory changes. Empirical experiments on large-scale real-world datasets from Citi Bike and Divvy demonstrate that SSGAN not only achieves state-of-the-art prediction accuracy but also significantly reduces operational costs compared to baseline models. This study provides a generalized, decision-oriented computerized methodology for optimizing BSS rebalancing operations.

共享单车系统（BSS）的指数级增长给供需管理带来了复杂的挑战，其中不平衡经常导致资源浪费和用户满意度降低。虽然图神经网络（gnn）已经成为需求预测的主流工具，但现有的方法主要依赖于静态地理邻近性，无法捕获由实际骑行行为驱动的潜在语义依赖性。为了弥补这一缺陷，本文提出了一种新的空间语义图注意神经网络（SSGAN）。与传统模型不同，SSGAN使用DTW构建语义邻接矩阵来量化站点流入和流出模式之间的形状相似性，从而捕获超越物理距离的非欧几里得相关性。此外，设计了一个门控多头注意机制，通过整合外部协变量（如天气）来动态权衡这些语义关系，使模型能够适应时变的上下文。至关重要的是，为了使预测准确性与决策有效性保持一致，该模型采用了双流架构，融合了流入和流出特征，以更好地反映净库存变化。来自Citi Bike和Divvy的大规模真实数据集的实证实验表明，与基线模型相比，SSGAN不仅达到了最先进的预测精度，而且显著降低了运营成本。本研究为优化BSS再平衡操作提供了一种通用的、决策导向的计算机化方法。

{"title":"Demand prediction for bike-sharing systems: A spatial and semantic modeling approach for enhanced accuracy and operational efficiency","authors":"Juntao Wu , Jiahui Feng , Jie Fang , Hefu Liu","doi":"10.1016/j.cie.2025.111775","DOIUrl":"10.1016/j.cie.2025.111775","url":null,"abstract":"<div><div>The exponential growth of Bike-Sharing Systems (BSS) has introduced complex challenges in supply–demand management, where imbalances frequently lead to resource wastage and reduced user satisfaction. While Graph Neural Networks (GNNs) have become a mainstream tool for demand forecasting, existing methodologies predominantly rely on static geographic proximity, failing to capture the latent semantic dependencies driven by actual riding behaviors. To bridge this gap, this paper proposes a novel Spatial-Semantic Graph Attention Neural Network (SSGAN). Unlike traditional models, SSGAN constructs a semantic adjacency matrix using DTW to quantify the shape similarity between station inflow and outflow patterns, thereby capturing non-Euclidean correlations beyond physical distance. Furthermore, a Gated Multi-Head Attention mechanism is designed to dynamically weigh these semantic relationships by integrating external covariates (e.g., weather), allowing the model to adapt to time-varying contexts. Crucially, to align prediction accuracy with decision effectiveness, the model employs a dual-stream architecture that fuses inflow and outflow features to better reflect net inventory changes. Empirical experiments on large-scale real-world datasets from Citi Bike and Divvy demonstrate that SSGAN not only achieves state-of-the-art prediction accuracy but also significantly reduces operational costs compared to baseline models. This study provides a generalized, decision-oriented computerized methodology for optimizing BSS rebalancing operations.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"213 ","pages":"Article 111775"},"PeriodicalIF":6.5,"publicationDate":"2025-12-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145841617","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