Computers in Industry最新文献_第7页

Leakage detection of oil and gas pipelines based on a multi-channel and multi-branch one-dimensional convolutional neural network with imbalanced samples 基于非平衡样本的多通道多分支一维卷积神经网络油气管道泄漏检测

IF 9.1 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers in Industry

Pub Date : 2025-09-08 DOI: 10.1016/j.compind.2025.104356

Dandi Yang , Peng Wang , Jingyi Lu , Chuang Guan , Hongli Dong

In recent years, intelligent pipeline leakage detection technology has played a crucial role in ensuring pipeline safety and energy security. However, most existing methods assume balanced datasets, overlooking the inherent imbalance between normal and abnormal data in real-world scenarios. This limitation hampers effective feature extraction for anomaly detection. To address this challenge, we propose a novel multi-channel and multi-branch one-dimensional convolutional neural network (MCB1DCNN). The model integrates a multi-channel convolution module and a multi-branch network structure to extract both global and local signal features. To mitigate the impact of data imbalance, we propose an adaptive weighted cross-entropy loss function. This function dynamically adjusts the loss weight of minority class samples based on the imbalance ratio. Furthermore, we construct a multi-channel acoustic signal dataset for oil and gas pipelines using the overlapping sample segmentation method. Variational mode decomposition (VMD) is applied to decompose acoustic signals into different frequency components, enabling comprehensive feature extraction. Ablation experiments analyze the impact of key model parameters. Experimental results show that MCB1DCNN outperforms several state-of-the-art methods in terms of accuracy, F1 score, false alarm rate, and missing alarm rate. These findings demonstrate its superior performance and practical applicability in real-world pipeline leakage detection.

近年来，智能管道泄漏检测技术在保障管道安全和能源安全方面发挥了至关重要的作用。然而，大多数现有方法假设数据集平衡，忽略了现实场景中正常和异常数据之间固有的不平衡。这一限制阻碍了异常检测中有效的特征提取。为了解决这一挑战，我们提出了一种新的多通道多分支一维卷积神经网络（MCB1DCNN）。该模型集成了多通道卷积模块和多分支网络结构，可同时提取全局和局部信号特征。为了减轻数据不平衡的影响，我们提出了一个自适应加权交叉熵损失函数。该函数根据失衡比例动态调整少数类样本的损失权重。在此基础上，利用重叠样本分割方法构建了油气管道多通道声信号数据集。采用变分模态分解（VMD）将声信号分解为不同的频率分量，从而实现全面的特征提取。烧蚀实验分析了关键模型参数的影响。实验结果表明，MCB1DCNN在准确率、F1分数、虚警率和漏警率方面都优于几种最先进的方法。这些结果表明了该方法在实际管道泄漏检测中的优越性能和实用性。

{"title":"Leakage detection of oil and gas pipelines based on a multi-channel and multi-branch one-dimensional convolutional neural network with imbalanced samples","authors":"Dandi Yang , Peng Wang , Jingyi Lu , Chuang Guan , Hongli Dong","doi":"10.1016/j.compind.2025.104356","DOIUrl":"10.1016/j.compind.2025.104356","url":null,"abstract":"<div><div>In recent years, intelligent pipeline leakage detection technology has played a crucial role in ensuring pipeline safety and energy security. However, most existing methods assume balanced datasets, overlooking the inherent imbalance between normal and abnormal data in real-world scenarios. This limitation hampers effective feature extraction for anomaly detection. To address this challenge, we propose a novel multi-channel and multi-branch one-dimensional convolutional neural network (MCB1DCNN). The model integrates a multi-channel convolution module and a multi-branch network structure to extract both global and local signal features. To mitigate the impact of data imbalance, we propose an adaptive weighted cross-entropy loss function. This function dynamically adjusts the loss weight of minority class samples based on the imbalance ratio. Furthermore, we construct a multi-channel acoustic signal dataset for oil and gas pipelines using the overlapping sample segmentation method. Variational mode decomposition (VMD) is applied to decompose acoustic signals into different frequency components, enabling comprehensive feature extraction. Ablation experiments analyze the impact of key model parameters. Experimental results show that MCB1DCNN outperforms several state-of-the-art methods in terms of accuracy, F1 score, false alarm rate, and missing alarm rate. These findings demonstrate its superior performance and practical applicability in real-world pipeline leakage detection.</div></div>","PeriodicalId":55219,"journal":{"name":"Computers in Industry","volume":"173 ","pages":"Article 104356"},"PeriodicalIF":9.1,"publicationDate":"2025-09-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145020142","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Data issues in industrial AI systems: A meta-review and research strategy 工业人工智能系统中的数据问题：元综述和研究策略

IF 9.1 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers in Industry

Pub Date : 2025-09-08 DOI: 10.1016/j.compind.2025.104361

Xuejiao Li , Yang Cheng , Charles Møller , Jay Lee

In the era of Industry 4.0, artificial intelligence (AI) is assumed to play an increasingly pivotal role within industrial systems. Despite the recent trend within various industries to adopt AI, the actual adoption of AI is not as developed as perceived. A significant factor contributing to this lag is the data issues in AI implementation. How to address these data issues stands as a significant concern confronting both industry and academia. Thus, this study conducts a comprehensive meta-review of data issues and corresponding methods in industrial AI. Eighty-two data issues are identified and categorized into seven stages of the data lifecycle. To supplement the existing research that focuses more on data issues arising in historical data, this study subsequently discusses the management of real-time sensor data and expert domain knowledge. Meanwhile, it proposes a model-aware data preparation approach, which integrates the data characteristics with specific AI model requirements to enhance data usability and algorithm alignment. This approach is further integrated into a conceptual framework that combines managerial and technical perspectives for systematically resolving data issues. The framework provides actionable insights and a systematic method for AI practitioners and industrial system developers to anticipate and address data-related challenges. Finally, the study highlights future research directions. This study advances the existing body of knowledge, supports a seamless transition from traditional model-centric AI to data-centric AI, and offers practical guidelines for professionals navigating the complexities of achieving data excellence in industrial AI applications.

在工业4.0时代，人工智能（AI）被认为在工业系统中发挥着越来越关键的作用。尽管最近各个行业都有采用人工智能的趋势，但人工智能的实际应用并不像人们想象的那样发达。造成这种滞后的一个重要因素是人工智能实施中的数据问题。如何解决这些数据问题是工业界和学术界面临的一个重大问题。因此，本研究对工业人工智能中的数据问题和相应方法进行了全面的元综述。确定了82个数据问题，并将其分为数据生命周期的七个阶段。为了补充现有研究更多地关注历史数据中出现的数据问题，本研究随后讨论了实时传感器数据和专家领域知识的管理。同时，提出了一种模型感知的数据准备方法，将数据特征与特定的AI模型需求相结合，增强数据可用性和算法一致性。这种方法进一步整合到一个概念框架中，该框架结合了系统地解决数据问题的管理和技术观点。该框架为人工智能从业者和工业系统开发人员提供了可操作的见解和系统方法，以预测和解决与数据相关的挑战。最后，对未来的研究方向进行了展望。这项研究推进了现有的知识体系，支持从传统的以模型为中心的人工智能向以数据为中心的人工智能的无缝过渡，并为专业人士提供了实用指南，帮助他们在工业人工智能应用中实现数据卓越的复杂性。

{"title":"Data issues in industrial AI systems: A meta-review and research strategy","authors":"Xuejiao Li , Yang Cheng , Charles Møller , Jay Lee","doi":"10.1016/j.compind.2025.104361","DOIUrl":"10.1016/j.compind.2025.104361","url":null,"abstract":"<div><div>In the era of Industry 4.0, artificial intelligence (AI) is assumed to play an increasingly pivotal role within industrial systems. Despite the recent trend within various industries to adopt AI, the actual adoption of AI is not as developed as perceived. A significant factor contributing to this lag is the data issues in AI implementation. How to address these data issues stands as a significant concern confronting both industry and academia. Thus, this study conducts a comprehensive meta-review of data issues and corresponding methods in industrial AI. Eighty-two data issues are identified and categorized into seven stages of the data lifecycle. To supplement the existing research that focuses more on data issues arising in historical data, this study subsequently discusses the management of real-time sensor data and expert domain knowledge. Meanwhile, it proposes a model-aware data preparation approach, which integrates the data characteristics with specific AI model requirements to enhance data usability and algorithm alignment. This approach is further integrated into a conceptual framework that combines managerial and technical perspectives for systematically resolving data issues. The framework provides actionable insights and a systematic method for AI practitioners and industrial system developers to anticipate and address data-related challenges. Finally, the study highlights future research directions. This study advances the existing body of knowledge, supports a seamless transition from traditional model-centric AI to data-centric AI, and offers practical guidelines for professionals navigating the complexities of achieving data excellence in industrial AI applications.</div></div>","PeriodicalId":55219,"journal":{"name":"Computers in Industry","volume":"173 ","pages":"Article 104361"},"PeriodicalIF":9.1,"publicationDate":"2025-09-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145020141","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","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 attention-free FCformer network with dynamic subdomain adaptation for open-set fault diagnosis of motor bearing 基于动态子域自适应的电机轴承开集故障诊断网络

IF 9.1 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers in Industry

Pub Date : 2025-09-04 DOI: 10.1016/j.compind.2025.104357

Chaoyang Weng, Baochun Lu, Longmiao Chen, Xiaoli Zhao, Wenbo Huang

Domain adaptation has emerged as an effective technique for addressing domain shift in motor bearing fault diagnosis. However, motor bearings often operate under harsh and uncertain conditions in actual industrial applications, where new fault modes may arise. This scenario gives rise to the open-set domain adaptation (OSDA) problem, which challenges the domain adaptation assumption that the source and target domains share the same label space. Therefore, this paper proposes a novel Attention-free FCformer network (AFCNet) with dynamic subdomain adaptation to address the OSDA fault diagnosis of motor bearing. Specifically, an improved Transformer based on Fourier and Convolutional embedding is first introduced to construct long-distance dependence in the frequency domain and extract more representative local domain-invariant fault features. Thereafter, an open-set subdomain adaptation module based on dynamic local maximum mean discrepancy is designed to align the conditional feature distribution of known classes by masking potential unknown classes. Furthermore, to reduce the impact of the empirical threshold setting on unknown class detection, an adaptive threshold learning (ATL) strategy is proposed to establish a reliable decision boundary between known and unknown classes. Finally, two fault diagnosis cases of motor bearing are carried out to validate the effectiveness and superiority of AFCNet. Experimental results demonstrate that AFCNet outperforms five benchmark models in terms of both accuracy and generalization across OSDA tasks with different source label spaces. These findings suggest that AFCNet offers a robust and reliable method for detecting new fault modes in motor bearings of rotating machinery.

领域自适应是解决电机轴承故障诊断中领域偏移问题的有效方法。然而，在实际工业应用中，电机轴承经常在恶劣和不确定的条件下运行，可能会出现新的故障模式。这种情况产生了开放集域自适应问题，它挑战了源域和目标域共享相同标签空间的域自适应假设。为此，本文提出了一种具有动态子域自适应的新型无注意力故障诊断网络（AFCNet）来解决电机轴承OSDA故障诊断问题。具体而言，首先引入了一种基于傅里叶和卷积嵌入的改进变压器，在频域上构造远距离依赖关系，提取更具代表性的局部域不变故障特征。然后，设计了一种基于动态局部最大均值差异的开集子域自适应模块，通过屏蔽潜在未知类来对齐已知类的条件特征分布。此外，为了减少经验阈值设置对未知类检测的影响，提出了自适应阈值学习（ATL）策略，在已知和未知类之间建立可靠的决策边界。最后，通过两个电机轴承故障诊断案例，验证了AFCNet的有效性和优越性。实验结果表明，在不同源标签空间的OSDA任务中，AFCNet在准确率和泛化方面都优于5个基准模型。这些发现表明，AFCNet为检测旋转机械电机轴承的新故障模式提供了一种鲁棒可靠的方法。

{"title":"A novel attention-free FCformer network with dynamic subdomain adaptation for open-set fault diagnosis of motor bearing","authors":"Chaoyang Weng, Baochun Lu, Longmiao Chen, Xiaoli Zhao, Wenbo Huang","doi":"10.1016/j.compind.2025.104357","DOIUrl":"10.1016/j.compind.2025.104357","url":null,"abstract":"<div><div>Domain adaptation has emerged as an effective technique for addressing domain shift in motor bearing fault diagnosis. However, motor bearings often operate under harsh and uncertain conditions in actual industrial applications, where new fault modes may arise. This scenario gives rise to the open-set domain adaptation (OSDA) problem, which challenges the domain adaptation assumption that the source and target domains share the same label space. Therefore, this paper proposes a novel Attention-free FCformer network (AFCNet) with dynamic subdomain adaptation to address the OSDA fault diagnosis of motor bearing. Specifically, an improved Transformer based on Fourier and Convolutional embedding is first introduced to construct long-distance dependence in the frequency domain and extract more representative local domain-invariant fault features. Thereafter, an open-set subdomain adaptation module based on dynamic local maximum mean discrepancy is designed to align the conditional feature distribution of known classes by masking potential unknown classes. Furthermore, to reduce the impact of the empirical threshold setting on unknown class detection, an adaptive threshold learning (ATL) strategy is proposed to establish a reliable decision boundary between known and unknown classes. Finally, two fault diagnosis cases of motor bearing are carried out to validate the effectiveness and superiority of AFCNet. Experimental results demonstrate that AFCNet outperforms five benchmark models in terms of both accuracy and generalization across OSDA tasks with different source label spaces. These findings suggest that AFCNet offers a robust and reliable method for detecting new fault modes in motor bearings of rotating machinery.</div></div>","PeriodicalId":55219,"journal":{"name":"Computers in Industry","volume":"173 ","pages":"Article 104357"},"PeriodicalIF":9.1,"publicationDate":"2025-09-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144989985","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

RLGBS: Reinforcement Learning-Guided Beam Search for process optimization in a paper machine dryer section 基于强化学习引导的束搜索的纸机干燥段工艺优化

IF 9.1 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers in Industry

Pub Date : 2025-08-28 DOI: 10.1016/j.compind.2025.104351

Siyuan Chen , Munevver Elif Asar , Jamal Yagoobi , Chenhui Shao

Paper drying is responsible for over two-thirds of energy consumption in the U.S. pulp and paper industry, presenting significant potential for energy savings through optimization of process parameters. Current approaches often assume fixed operating conditions, neglecting dynamic ambient and process variations that limit achievable savings and real-world applicability. To this end, we develop a physics-based simulation environment for a paper machine dryer section and propose a reinforcement learning (RL) framework to minimize overall energy consumption by optimizing drying process parameters under diverse operating conditions. To mitigate overdrying and numerical instabilities caused by suboptimal local RL actions, we introduce Reinforcement Learning-Guided Beam Search (RLGBS), which explores multiple action sequences in parallel using beam search. Instead of making step-by-step decisions, RLGBS prioritizes solutions based on cumulative probability, reducing the impact of individual suboptimal actions. Experiments demonstrate that RLGBS achieves consistent energy savings under unseen operating conditions not encountered during training, outperforming conventional RL methods. While validated in drying optimization, this framework is broadly applicable to other RL-based industrial process control problems.

在美国纸浆和造纸工业中，纸张干燥占能源消耗的三分之二以上，通过优化工艺参数，呈现出巨大的节能潜力。目前的方法通常假设固定的操作条件，忽略了动态环境和工艺变化，这些变化限制了可实现的节约和实际应用。为此，我们开发了一个基于物理的纸机干燥段仿真环境，并提出了一个强化学习（RL）框架，通过优化不同操作条件下的干燥过程参数来最大限度地降低总体能耗。为了减轻由次优局部RL动作引起的过度干燥和数值不稳定，我们引入了强化学习引导束搜索（RLGBS），它使用束搜索并行地探索多个动作序列。RLGBS不是一步一步地做决定，而是根据累积概率来确定解决方案的优先级，从而减少了单个次优行为的影响。实验表明，RLGBS在训练过程中未遇到的未知操作条件下实现了一致的节能，优于传统的RL方法。虽然在干燥优化中得到了验证，但该框架广泛适用于其他基于rl的工业过程控制问题。

{"title":"RLGBS: Reinforcement Learning-Guided Beam Search for process optimization in a paper machine dryer section","authors":"Siyuan Chen , Munevver Elif Asar , Jamal Yagoobi , Chenhui Shao","doi":"10.1016/j.compind.2025.104351","DOIUrl":"10.1016/j.compind.2025.104351","url":null,"abstract":"<div><div>Paper drying is responsible for over two-thirds of energy consumption in the U.S. pulp and paper industry, presenting significant potential for energy savings through optimization of process parameters. Current approaches often assume fixed operating conditions, neglecting dynamic ambient and process variations that limit achievable savings and real-world applicability. To this end, we develop a physics-based simulation environment for a paper machine dryer section and propose a reinforcement learning (RL) framework to minimize overall energy consumption by optimizing drying process parameters under diverse operating conditions. To mitigate overdrying and numerical instabilities caused by suboptimal local RL actions, we introduce Reinforcement Learning-Guided Beam Search (RLGBS), which explores multiple action sequences in parallel using beam search. Instead of making step-by-step decisions, RLGBS prioritizes solutions based on cumulative probability, reducing the impact of individual suboptimal actions. Experiments demonstrate that RLGBS achieves consistent energy savings under unseen operating conditions not encountered during training, outperforming conventional RL methods. While validated in drying optimization, this framework is broadly applicable to other RL-based industrial process control problems.</div></div>","PeriodicalId":55219,"journal":{"name":"Computers in Industry","volume":"172 ","pages":"Article 104351"},"PeriodicalIF":9.1,"publicationDate":"2025-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144908310","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Knowledge graph construction with meta-learning for continuously accumulated manufacturing knowledge 利用元学习构建知识图谱，实现制造知识的持续积累

IF 9.1 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers in Industry

Pub Date : 2025-08-28 DOI: 10.1016/j.compind.2025.104353

Yanzhen Jing , Guanghui Zhou , Chao Zhang , Fengtian Chang , Jiacheng Li

The construction of manufacturing knowledge graph (MKG) has been regarded as an important technical roadmap to support designer-oriented manufacturing knowledge reuse. It can improve product manufacturability and reduce design iterations. However, manufacturing knowledge is lesson-learned texts of enterprises. Traditional deep learning-driven MKG construction requires sufficient training samples, which heavily rely on manual labeling. It is both time-consuming and labor-intensive. Meanwhile, due to the new manufacturing knowledge accumulation, an MKG also needs to be continuously updated. To bridge the gap, this paper proposes an efficient MKG construction approach with meta-learning. Based on the manufacturing knowledge ontology, a novel two-stage knowledge extraction model (TKEM) is presented to achieve low-resource entity recognition. Then, considering the newly accumulated manufacturing knowledge, a continuous knowledge fusion strategy is illustrated to complete the MKG construction and update. Finally, the experimental results show that the TKEM outperforms state-of-the-art baselines on both the manufacturing knowledge dataset and a public dataset. In addition, a prototype system provides the application of MKG-based manufacturing knowledge reuse, which can perceive explicit and implicit knowledge requirements of designers by MKG embedding learning.

制造知识图谱的构建已被视为支持面向设计人员的制造知识复用的重要技术路线。它可以提高产品的可制造性，减少设计迭代。然而，制造知识是企业的经验教训文本。传统的深度学习驱动MKG构建需要足够的训练样本，严重依赖人工标注。这既费时又费力。同时，由于新的制造知识积累，MKG也需要不断更新。为了弥补这一差距，本文提出了一种基于元学习的高效MKG构建方法。在制造知识本体的基础上，提出了一种新的两阶段知识抽取模型（TKEM）来实现低资源实体识别。然后，考虑到新积累的制造知识，提出了一种持续的知识融合策略来完成MKG的构建和更新。最后，实验结果表明，TKEM在制造知识数据集和公共数据集上都优于最先进的基线。此外，一个原型系统提供了基于MKG的制造知识复用应用，该系统可以通过MKG嵌入学习感知设计者的显性和隐性知识需求。

{"title":"Knowledge graph construction with meta-learning for continuously accumulated manufacturing knowledge","authors":"Yanzhen Jing , Guanghui Zhou , Chao Zhang , Fengtian Chang , Jiacheng Li","doi":"10.1016/j.compind.2025.104353","DOIUrl":"10.1016/j.compind.2025.104353","url":null,"abstract":"<div><div>The construction of manufacturing knowledge graph (MKG) has been regarded as an important technical roadmap to support designer-oriented manufacturing knowledge reuse. It can improve product manufacturability and reduce design iterations. However, manufacturing knowledge is lesson-learned texts of enterprises. Traditional deep learning-driven MKG construction requires sufficient training samples, which heavily rely on manual labeling. It is both time-consuming and labor-intensive. Meanwhile, due to the new manufacturing knowledge accumulation, an MKG also needs to be continuously updated. To bridge the gap, this paper proposes an efficient MKG construction approach with meta-learning. Based on the manufacturing knowledge ontology, a novel two-stage knowledge extraction model (TKEM) is presented to achieve low-resource entity recognition. Then, considering the newly accumulated manufacturing knowledge, a continuous knowledge fusion strategy is illustrated to complete the MKG construction and update. Finally, the experimental results show that the TKEM outperforms state-of-the-art baselines on both the manufacturing knowledge dataset and a public dataset. In addition, a prototype system provides the application of MKG-based manufacturing knowledge reuse, which can perceive explicit and implicit knowledge requirements of designers by MKG embedding learning.</div></div>","PeriodicalId":55219,"journal":{"name":"Computers in Industry","volume":"172 ","pages":"Article 104353"},"PeriodicalIF":9.1,"publicationDate":"2025-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144912916","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Carbon footprint optimization of a LoRa-based multi-hop Industrial Internet of Things network deployment 基于lora的多跳工业物联网网络部署的碳足迹优化

IF 9.1 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers in Industry

Pub Date : 2025-08-27 DOI: 10.1016/j.compind.2025.104348

Francisco-Jose Alvarado-Alcon, Rafael Asorey-Cacheda, Antonio-Javier Garcia-Sanchez, Joan Garcia-Haro

The seamless integration of the Internet of Things (IoT) into various societal and economic domains is unfolding before us. In the industrial sector, this digital transformation propelled by Industrial Internet of Things (IIoT) networks, among other technologies is referred to as Industry 4.0, where wireless technologies are transforming the industry as currently conceived. While considerable efforts have been devoted to optimizing performance and energy consumption in these networks, relatively little attention has been directed towards comprehensively studying and optimizing the carbon footprint (CF) associated with these network deployments. Additionally, scarce literature has analyzed the use of multi-hop topologies with well-known standards like LoRaWAN. This research delves into the CF of a generic multi-hop IIoT network using renewable energy sources and communicating through the LoRa physical layer while also proposing an optimization framework. The findings indicate that up to a 85% reduction in carbon emissions can be achieved by enabling packet forwarding through end devices, offering greater scalability. Interestingly, an optimal end device density emerges, implying that decreasing the number of end devices may actually lead to higher CFs. These results underscore the necessity for a fresh perspective on optimizing IIoT networks, urging the inclusion of environmental sustainability criteria that have hitherto been overlooked.

物联网（IoT）与各种社会和经济领域的无缝集成正在我们面前展开。在工业领域，这种由工业物联网（IIoT）网络推动的数字化转型，以及其他技术被称为工业4.0，其中无线技术正在改变目前设想的行业。虽然在优化这些网络的性能和能源消耗方面已经投入了相当大的努力，但对与这些网络部署相关的碳足迹（CF）进行全面研究和优化的关注相对较少。此外，很少有文献分析了LoRaWAN等知名标准中多跳拓扑的使用。本研究深入研究了使用可再生能源并通过LoRa物理层进行通信的通用多跳IIoT网络的CF，同时提出了优化框架。研究结果表明，通过启用终端设备的数据包转发，可以减少高达85%的碳排放，从而提供更大的可扩展性。有趣的是，出现了一个最佳的终端设备密度，这意味着减少终端设备的数量实际上可能导致更高的CFs。这些结果强调了优化工业物联网网络的新视角的必要性，敦促将迄今为止被忽视的环境可持续性标准纳入其中。

{"title":"Carbon footprint optimization of a LoRa-based multi-hop Industrial Internet of Things network deployment","authors":"Francisco-Jose Alvarado-Alcon, Rafael Asorey-Cacheda, Antonio-Javier Garcia-Sanchez, Joan Garcia-Haro","doi":"10.1016/j.compind.2025.104348","DOIUrl":"10.1016/j.compind.2025.104348","url":null,"abstract":"<div><div>The seamless integration of the Internet of Things (IoT) into various societal and economic domains is unfolding before us. In the industrial sector, this digital transformation propelled by Industrial Internet of Things (IIoT) networks, among other technologies is referred to as Industry 4.0, where wireless technologies are transforming the industry as currently conceived. While considerable efforts have been devoted to optimizing performance and energy consumption in these networks, relatively little attention has been directed towards comprehensively studying and optimizing the carbon footprint (CF) associated with these network deployments. Additionally, scarce literature has analyzed the use of multi-hop topologies with well-known standards like LoRaWAN. This research delves into the CF of a generic multi-hop IIoT network using renewable energy sources and communicating through the LoRa physical layer while also proposing an optimization framework. The findings indicate that up to a 85% reduction in carbon emissions can be achieved by enabling packet forwarding through end devices, offering greater scalability. Interestingly, an optimal end device density emerges, implying that decreasing the number of end devices may actually lead to higher CFs. These results underscore the necessity for a fresh perspective on optimizing IIoT networks, urging the inclusion of environmental sustainability criteria that have hitherto been overlooked.</div></div>","PeriodicalId":55219,"journal":{"name":"Computers in Industry","volume":"172 ","pages":"Article 104348"},"PeriodicalIF":9.1,"publicationDate":"2025-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144908309","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Semantic middleware for demand response systems: Enhancing data interoperability in green electricity management for manufacturing 需求响应系统的语义中间件：增强制造业绿色电力管理中的数据互操作性

IF 9.1 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers in Industry

Pub Date : 2025-08-27 DOI: 10.1016/j.compind.2025.104354

Tina Boroukhian , Kritkorn Supyen , Christopher William Mclaughlan , Atit Bashyal , Tuan Pham , Hendro Wicaksono

Optimizing the consumption of green electricity across sectors, including manufacturing, is a critical strategy for achieving net-zero emissions and advancing clean production in Europe by 2050. Demand Response (DR) represents a promising approach to engaging power consumers from all sectors in the transition toward increased utilization of renewable energy sources. A functional DR system for manufacturing power consumers requires seamless data integration and communication between information systems across multiple domains, including both power consumption and generation. This paper introduces a semantic middleware specifically designed for DR systems in the manufacturing sector, using an ontology as the central information model. To develop this ontology, we adopted a strategy that reuses and unifies existing ontologies from multiple domains, ensuring comprehensive coverage of the data requirements for DR applications in manufacturing. To operationalize this strategy, we designed novel methods for effective ontology unification and implemented them within a dedicated unification tool. This process was followed by data-to-ontology mapping to construct a knowledge graph, and was further extended through the development of a querying system equipped with a natural language interface. Additionally, this paper offers insights into the deployment environment of the semantic middleware, encompassing multiple data sources and the applications that utilize this data. The proposed approach is implemented in multiple German manufacturing small and medium-sized enterprises connected to a utility company, demonstrating consistent data interpretation and seamless information integration. Consequently, the method offers practical potential for optimizing green electricity usage in the manufacturing sector and supporting the transition toward a more sustainable and cleaner future.

优化包括制造业在内的各个部门的绿色电力消费，是到2050年实现欧洲净零排放和推进清洁生产的关键战略。需求响应（DR）代表了一种很有前途的方法，可以让所有部门的电力消费者参与到增加可再生能源利用的过渡中。针对制造业用电量用户的容灾系统需要跨多域信息系统之间的无缝数据集成和通信，包括功耗和发电。本文采用本体作为中心信息模型，介绍了一种专门为制造业的灾难恢复系统设计的语义中间件。为了开发这个本体，我们采用了一种策略，重用和统一来自多个领域的现有本体，确保全面覆盖制造业中DR应用程序的数据需求。为了实现这一策略，我们设计了新的方法来实现有效的本体统一，并在专用的统一工具中实现它们。在此过程之后，通过数据到本体的映射来构建知识图谱，并通过开发配备自然语言接口的查询系统进一步扩展。此外，本文还深入介绍了语义中间件的部署环境，包括多个数据源和利用这些数据的应用程序。所提出的方法在与公用事业公司相连的多个德国制造业中小型企业中实施，展示了一致的数据解释和无缝的信息集成。因此，该方法为优化制造业的绿色电力使用提供了实际潜力，并支持向更可持续和更清洁的未来过渡。

{"title":"Semantic middleware for demand response systems: Enhancing data interoperability in green electricity management for manufacturing","authors":"Tina Boroukhian , Kritkorn Supyen , Christopher William Mclaughlan , Atit Bashyal , Tuan Pham , Hendro Wicaksono","doi":"10.1016/j.compind.2025.104354","DOIUrl":"10.1016/j.compind.2025.104354","url":null,"abstract":"<div><div>Optimizing the consumption of green electricity across sectors, including manufacturing, is a critical strategy for achieving net-zero emissions and advancing clean production in Europe by 2050. Demand Response (DR) represents a promising approach to engaging power consumers from all sectors in the transition toward increased utilization of renewable energy sources. A functional DR system for manufacturing power consumers requires seamless data integration and communication between information systems across multiple domains, including both power consumption and generation. This paper introduces a semantic middleware specifically designed for DR systems in the manufacturing sector, using an ontology as the central information model. To develop this ontology, we adopted a strategy that reuses and unifies existing ontologies from multiple domains, ensuring comprehensive coverage of the data requirements for DR applications in manufacturing. To operationalize this strategy, we designed novel methods for effective ontology unification and implemented them within a dedicated unification tool. This process was followed by data-to-ontology mapping to construct a knowledge graph, and was further extended through the development of a querying system equipped with a natural language interface. Additionally, this paper offers insights into the deployment environment of the semantic middleware, encompassing multiple data sources and the applications that utilize this data. The proposed approach is implemented in multiple German manufacturing small and medium-sized enterprises connected to a utility company, demonstrating consistent data interpretation and seamless information integration. Consequently, the method offers practical potential for optimizing green electricity usage in the manufacturing sector and supporting the transition toward a more sustainable and cleaner future.</div></div>","PeriodicalId":55219,"journal":{"name":"Computers in Industry","volume":"172 ","pages":"Article 104354"},"PeriodicalIF":9.1,"publicationDate":"2025-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144908308","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","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 interval-valued carbon price forecasting paradigm: multi-factor intelligent recognition-based ensemble learning 一种新的区间碳价预测范式：基于多因素智能识别的集成学习

IF 9.1 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers in Industry

Pub Date : 2025-08-25 DOI: 10.1016/j.compind.2025.104352

Yan Hao , Xiaodi Wang , Wendong Yang

Accurate carbon price prediction is of significant importance for the sustainable growth of the carbon market, as it influences the realization of dual carbon goals and the transition toward a low-carbon economy. However, previous forecasting models have typically been point-valued carbon price-based, and the factors influencing carbon prices have not been comprehensively identified. Therefore, this study proposes a novel multi-factor intelligent recognition-based ensemble forecasting system for interval-valued carbon price forecasts. In this system, factors from multiple perspectives are considered to analyze interval-valued carbon price fluctuations. To select the optimal set of influencing factors, a multi-factor intelligent recognition subsystem combining a time-series causal analysis method with multi-objective feature selection algorithms was developed. This subsystem simultaneously considers the intrinsic correlations among factors and the predictive performance to thereby ensure the accuracy of feature selection while reducing redundancy. Additionally, an ensemble forecasting subsystem integrating multiple machine learning models was constructed to exploit the merits of each model and realize more accurate results than can be achieved by any individual model. Empirical research demonstrated that this forecasting system could accurately identify powerful influencing factors, outperform other feature selection strategies, and achieve interval-valued mean absolute percentage errors of 1.5883 % and 1.5113 %, respectively. Therefore, this system is an effective tool for predicting carbon prices.

准确的碳价格预测对碳市场的可持续发展具有重要意义，因为它影响到双碳目标的实现和向低碳经济的转型。然而，以前的预测模型通常是基于点价值碳价格的，并且没有全面确定影响碳价格的因素。为此，本文提出了一种基于多因素智能识别的区间碳价集成预测系统。在这个系统中，考虑了多个角度的因素来分析区间价值碳价格波动。为了选择最优的影响因素集，开发了将时间序列因果分析方法与多目标特征选择算法相结合的多因素智能识别子系统。该子系统同时考虑了因素之间的内在相关性和预测性能，从而在减少冗余的同时保证了特征选择的准确性。此外，构建了一个集成多个机器学习模型的集成预测子系统，以利用每个模型的优点，实现比单个模型更准确的预测结果。实证研究表明，该预测系统能够准确识别强大的影响因素，优于其他特征选择策略，区间均值绝对百分比误差分别为1.5883 %和1.5113 %。因此，该系统是预测碳价格的有效工具。

{"title":"A novel interval-valued carbon price forecasting paradigm: multi-factor intelligent recognition-based ensemble learning","authors":"Yan Hao , Xiaodi Wang , Wendong Yang","doi":"10.1016/j.compind.2025.104352","DOIUrl":"10.1016/j.compind.2025.104352","url":null,"abstract":"<div><div>Accurate carbon price prediction is of significant importance for the sustainable growth of the carbon market, as it influences the realization of dual carbon goals and the transition toward a low-carbon economy. However, previous forecasting models have typically been point-valued carbon price-based, and the factors influencing carbon prices have not been comprehensively identified. Therefore, this study proposes a novel multi-factor intelligent recognition-based ensemble forecasting system for interval-valued carbon price forecasts. In this system, factors from multiple perspectives are considered to analyze interval-valued carbon price fluctuations. To select the optimal set of influencing factors, a multi-factor intelligent recognition subsystem combining a time-series causal analysis method with multi-objective feature selection algorithms was developed. This subsystem simultaneously considers the intrinsic correlations among factors and the predictive performance to thereby ensure the accuracy of feature selection while reducing redundancy. Additionally, an ensemble forecasting subsystem integrating multiple machine learning models was constructed to exploit the merits of each model and realize more accurate results than can be achieved by any individual model. Empirical research demonstrated that this forecasting system could accurately identify powerful influencing factors, outperform other feature selection strategies, and achieve interval-valued mean absolute percentage errors of 1.5883 % and 1.5113 %, respectively. Therefore, this system is an effective tool for predicting carbon prices.</div></div>","PeriodicalId":55219,"journal":{"name":"Computers in Industry","volume":"172 ","pages":"Article 104352"},"PeriodicalIF":9.1,"publicationDate":"2025-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144896712","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

An agent-based approach for the automatic generation of valid SysMLv2 Models in industrial contexts 用于在工业环境中自动生成有效SysMLv2模型的基于代理的方法

IF 9.1 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers in Industry

Pub Date : 2025-08-24 DOI: 10.1016/j.compind.2025.104350

Eduardo Cibrián, Jose Olivert-Iserte, Juan Llorens, Jose María Álvarez-Rodríguez

Automating the generation of valid SysML v2 models from natural language specifications holds promise for advancing Model-Based Systems Engineering (MBSE) in industrial settings. However, current approaches based solely on Large Language Models (LLMs) often fail to meet the syntactic and semantic rigor required by formal modeling languages. This paper introduces a domain-informed, agent-based framework that combines LLMs with structured retrieval and iterative validation to synthesize correct SysML v2 models. The system integrates Retrieval-Augmented Generation (RAG) using a curated repository of SysML v2 examples and enforces compliance through a validation engine based on the official ANTLR grammar. Experimental results across diverse MBSE scenarios demonstrate that the integration of retrieval and validation mechanisms leads to a substantial improvement in model correctness and semantic alignment, beyond what each component achieves individually. This combined effect enables reliable, closed-loop generation of formal models from natural language, illustrating how domain-specific integration can transform general-purpose LLMs into reliable assistants for engineering design tasks.

从自然语言规范中自动生成有效的SysML v2模型，有望在工业环境中推进基于模型的系统工程（MBSE）。然而，目前仅基于大型语言模型（llm）的方法往往不能满足正式建模语言所要求的语法和语义严密性。本文介绍了一个领域知情的、基于代理的框架，该框架将llm与结构化检索和迭代验证相结合，以合成正确的SysML v2模型。该系统使用SysML v2示例的管理存储库集成了检索增强生成（RAG），并通过基于官方ANTLR语法的验证引擎强制执行遵从性。跨不同MBSE场景的实验结果表明，检索和验证机制的集成大大提高了模型正确性和语义一致性，超出了每个组件单独实现的范围。这种组合效应使得从自然语言生成可靠的、闭环的正式模型成为可能，说明了特定领域的集成如何将通用llm转换为工程设计任务的可靠助手。

{"title":"An agent-based approach for the automatic generation of valid SysMLv2 Models in industrial contexts","authors":"Eduardo Cibrián, Jose Olivert-Iserte, Juan Llorens, Jose María Álvarez-Rodríguez","doi":"10.1016/j.compind.2025.104350","DOIUrl":"10.1016/j.compind.2025.104350","url":null,"abstract":"<div><div>Automating the generation of valid SysML v2 models from natural language specifications holds promise for advancing Model-Based Systems Engineering (MBSE) in industrial settings. However, current approaches based solely on Large Language Models (LLMs) often fail to meet the syntactic and semantic rigor required by formal modeling languages. This paper introduces a domain-informed, agent-based framework that combines LLMs with structured retrieval and iterative validation to synthesize correct SysML v2 models. The system integrates Retrieval-Augmented Generation (RAG) using a curated repository of SysML v2 examples and enforces compliance through a validation engine based on the official ANTLR grammar. Experimental results across diverse MBSE scenarios demonstrate that the integration of retrieval and validation mechanisms leads to a substantial improvement in model correctness and semantic alignment, beyond what each component achieves individually. This combined effect enables reliable, closed-loop generation of formal models from natural language, illustrating how domain-specific integration can transform general-purpose LLMs into reliable assistants for engineering design tasks.</div></div>","PeriodicalId":55219,"journal":{"name":"Computers in Industry","volume":"172 ","pages":"Article 104350"},"PeriodicalIF":9.1,"publicationDate":"2025-08-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144892925","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

An online milling deformation prediction method for thin-walled features with domain adversarial neural networks under small samples 基于领域对抗神经网络的小样本薄壁特征铣削变形在线预测方法

IF 9.1 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers in Industry

Pub Date : 2025-08-18 DOI: 10.1016/j.compind.2025.104349

Jingjing Li , Guanghui Zhou , Chao Zhang , Zhijie Wei , Fengyi Lu

Thin-walled machining features are extensively utilized in the aerospace industry, where the milling deformation caused by their weak rigidity has been the most common quality concern. Efficient control of milling deformation for thin-walled features is essential for enhancing quality. However, the high cost and time-consuming nature of data collection for aviation parts, leading to a limited availability of process data, which presents a significant challenge for predicting deformation in aerospace components. To address this issue, this study aims to develop a high-precision milling deformation prediction method by fully leveraging the small-sample data from machining experiments and simulation data. This paper first constructs a thin-walled features deformation prediction framework by integrating Domain Adversarial Neural Networks (DANN) with a digital twin process model. Secondly, the DANN method is adopted to achieve online prediction of milling deformation for thin-walled features. A small quantity of experimental deformation data serves as the target domain for training dataset, whereas milling simulation data produced by finite element software serves as the source domain. Milling deformation is accurately predicted using adversarial training based on the DANN structure for domain regression and domain classification. The best results show that the proposed method achieves better goodness of fit under limited sample conditions, with a 5 % increase in the Coefficient of Determination (R²) and a 15 % reduction in Mean Absolute Error (MAE) compared to five baseline methods. In the end, the DANN approach was integrated into the digital twin system for the milling process, and a prototype system was constructed to demonstrate the viability of the suggested approach.

薄壁加工特性在航空航天工业中得到了广泛的应用，由于薄壁加工的弱刚性引起的铣削变形一直是航空航天工业中最常见的质量问题。薄壁件铣削变形的有效控制是提高加工质量的关键。然而，航空部件数据收集的高成本和耗时特性导致过程数据的可用性有限，这对预测航空部件的变形提出了重大挑战。为了解决这一问题，本研究旨在充分利用加工实验和仿真数据的小样本数据，开发高精度铣削变形预测方法。本文首先将领域对抗神经网络（DANN）与数字孪生过程模型相结合，构建了薄壁特征变形预测框架。其次，采用DANN方法实现了薄壁特征铣削变形的在线预测；少量的实验变形数据作为训练数据集的目标域，而有限元软件生成的铣削仿真数据作为源域。利用基于DANN结构的对抗训练进行领域回归和领域分类，准确预测铣削变形。最佳结果表明，该方法在有限样本条件下获得了更好的拟合优度，与五种基线方法相比，决定系数（R²）提高了5 %，平均绝对误差（MAE）降低了15 %。最后，将DANN方法集成到铣削过程的数字孪生系统中，并构建了一个原型系统来验证所建议方法的可行性。

{"title":"An online milling deformation prediction method for thin-walled features with domain adversarial neural networks under small samples","authors":"Jingjing Li , Guanghui Zhou , Chao Zhang , Zhijie Wei , Fengyi Lu","doi":"10.1016/j.compind.2025.104349","DOIUrl":"10.1016/j.compind.2025.104349","url":null,"abstract":"<div><div>Thin-walled machining features are extensively utilized in the aerospace industry, where the milling deformation caused by their weak rigidity has been the most common quality concern. Efficient control of milling deformation for thin-walled features is essential for enhancing quality. However, the high cost and time-consuming nature of data collection for aviation parts, leading to a limited availability of process data, which presents a significant challenge for predicting deformation in aerospace components. To address this issue, this study aims to develop a high-precision milling deformation prediction method by fully leveraging the small-sample data from machining experiments and simulation data. This paper first constructs a thin-walled features deformation prediction framework by integrating Domain Adversarial Neural Networks (DANN) with a digital twin process model. Secondly, the DANN method is adopted to achieve online prediction of milling deformation for thin-walled features. A small quantity of experimental deformation data serves as the target domain for training dataset, whereas milling simulation data produced by finite element software serves as the source domain. Milling deformation is accurately predicted using adversarial training based on the DANN structure for domain regression and domain classification. The best results show that the proposed method achieves better goodness of fit under limited sample conditions, with a 5 % increase in the Coefficient of Determination (R²) and a 15 % reduction in Mean Absolute Error (MAE) compared to five baseline methods. In the end, the DANN approach was integrated into the digital twin system for the milling process, and a prototype system was constructed to demonstrate the viability of the suggested approach.</div></div>","PeriodicalId":55219,"journal":{"name":"Computers in Industry","volume":"172 ","pages":"Article 104349"},"PeriodicalIF":9.1,"publicationDate":"2025-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144865067","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0