Journal of Industrial Information Integration最新文献_第8页

CoperFed: A covert personalized federated learning framework for Industrial Control Systems intrusion detection 用于工业控制系统入侵检测的隐蔽个性化联邦学习框架

IF 10.4 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Journal of Industrial Information Integration

Pub Date : 2025-11-01 DOI: 10.1016/j.jii.2025.101004

Yao Shan, Jindong Zhao, Yongchao Song, Haojun Teng, Wenhan Hou, Zhaowei Liu

Modern information and communication technologies have propelled transformative modernization of Industrial Control Systems (ICSs) while exacerbating cybersecurity risks. Federated Learning (FL) offers a privacy-preserving framework for collaborative development of intrusion detection models among distributed participants. However, its effectiveness is significantly limited by inherent model divergence caused by non-independent and identically distributed (Non-IID) data characteristics. Moreover, direct implementation of FL in ICS environments faces critical challenges due to insufficient capabilities in network traffic feature representation and device concealment. To address these challenges, we propose CoperFed, a covert personalized FL framework that generates unique intrusion detection models for individual participants. First, we developed Gicsmeter, a multi-dimensional ICS traffic representation tool for all participants, to enhance model performance at the data level. Second, we designed a personalized update algorithm based on key model parameters to improve collaboration among similar participants. By integrating global knowledge during model aggregation, this algorithm equips the model with local and global scenario detection capabilities. Finally, we designed a covert federated communication scheme for ICS that can effectively conceal the federated training process within regular ICS traffic and reduce the exposure risk of FL participants. Experiments show that CoperFed outperforms baseline methods in intrusion detection and robustness and can effectively divert attackers’ attention from FL participants.

现代信息和通信技术推动了工业控制系统（ics）的变革性现代化，同时加剧了网络安全风险。联邦学习（FL）为分布式参与者之间协作开发入侵检测模型提供了一个隐私保护框架。然而，非独立和同分布（Non-IID）数据特征导致的固有模型发散严重限制了其有效性。此外，由于网络流量特征表示和设备隐藏能力不足，在ICS环境中直接实现FL面临着严峻的挑战。为了解决这些挑战，我们提出了cooperfed，这是一个隐蔽的个性化FL框架，可以为个体参与者生成独特的入侵检测模型。首先，我们为所有参与者开发了一个多维ICS流量表示工具Gicsmeter，以提高模型在数据层面的性能。其次，设计了基于关键模型参数的个性化更新算法，以提高相似参与者之间的协作能力。该算法通过在模型聚合过程中集成全局知识，使模型具有局部和全局场景检测能力。最后，我们为ICS设计了一种隐蔽的联邦通信方案，可以有效地将联邦训练过程隐藏在常规ICS流量中，降低FL参与者的暴露风险。实验表明，cooperfed在入侵检测和鲁棒性方面优于基线方法，能够有效转移攻击者对FL参与者的注意力。

{"title":"CoperFed: A covert personalized federated learning framework for Industrial Control Systems intrusion detection","authors":"Yao Shan, Jindong Zhao, Yongchao Song, Haojun Teng, Wenhan Hou, Zhaowei Liu","doi":"10.1016/j.jii.2025.101004","DOIUrl":"10.1016/j.jii.2025.101004","url":null,"abstract":"<div><div>Modern information and communication technologies have propelled transformative modernization of Industrial Control Systems (ICSs) while exacerbating cybersecurity risks. Federated Learning (FL) offers a privacy-preserving framework for collaborative development of intrusion detection models among distributed participants. However, its effectiveness is significantly limited by inherent model divergence caused by non-independent and identically distributed (Non-IID) data characteristics. Moreover, direct implementation of FL in ICS environments faces critical challenges due to insufficient capabilities in network traffic feature representation and device concealment. To address these challenges, we propose CoperFed, a covert personalized FL framework that generates unique intrusion detection models for individual participants. First, we developed Gicsmeter, a multi-dimensional ICS traffic representation tool for all participants, to enhance model performance at the data level. Second, we designed a personalized update algorithm based on key model parameters to improve collaboration among similar participants. By integrating global knowledge during model aggregation, this algorithm equips the model with local and global scenario detection capabilities. Finally, we designed a covert federated communication scheme for ICS that can effectively conceal the federated training process within regular ICS traffic and reduce the exposure risk of FL participants. Experiments show that CoperFed outperforms baseline methods in intrusion detection and robustness and can effectively divert attackers’ attention from FL participants.</div></div>","PeriodicalId":55975,"journal":{"name":"Journal of Industrial Information Integration","volume":"48 ","pages":"Article 101004"},"PeriodicalIF":10.4,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145461741","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 EMD-based forecasting framework integrating GMM and BiLSTM for helicopter engine anomaly detection 结合GMM和BiLSTM的直升机发动机异常检测emd预测框架

IF 10.4 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Journal of Industrial Information Integration

Pub Date : 2025-11-01 DOI: 10.1016/j.jii.2025.101003

Qi Shen , Jingwei Guo , Yihui Tian , Zhen-Song Chen

The safety of helicopter operations is paramount, yet early signs of potential failures often go undetected, highlighting the need for robust signal alert systems during flights. Detecting anomalies in helicopter engine behavior through vibration analysis is critically important due to the long-sequence nature and complexity of the data, which present significant challenges for real-time assessment and are not adequately addressed by traditional methods such as preset thresholds or basic statistical models, as these approaches struggle to capture intricate spatiotemporal dependencies and overlapping fault patterns in real-world scenarios. To address these challenges, we introduce a novel hybrid model that leverages Empirical Mode Decomposition (EMD) for signal decomposition and analysis, effectively overcoming the limitations of traditional approaches. EMD is particularly advantageous as it decomposes complex signals into Intrinsic Mode Functions (IMFs), enabling more accurate anomaly detection in long sequences. Following EMD, the Gaussian Mixture Model (GMM) is employed to precisely recognize various fault patterns, ensuring a robust foundation for anomaly detection. Bidirectional Long Short-Term Memory (BiLSTM) networks further enhance the model by capturing temporal dependencies in both directions, integrating critical spatiotemporal information and improving predictive accuracy. Experimental results demonstrate that this integrated EMD-GMM-BiLSTM approach is not only highly sensitive and accurate in detecting anomalies but also significantly simpler and more efficient than more complex frameworks such as encoder-decoder models or Transformers. This method ensures the operational safety of helicopters and supports the broader adoption of low-altitude economic activities by providing essential safety guarantees.

直升机操作的安全是至关重要的，然而潜在故障的早期迹象往往未被发现，这突出了在飞行过程中对强大的信号警报系统的需求。由于数据的长序列性质和复杂性，通过振动分析检测直升机发动机异常行为至关重要，这对实时评估提出了重大挑战，并且无法通过预设阈值或基本统计模型等传统方法充分解决，因为这些方法难以捕捉复杂的时空依赖性和重叠故障模式。为了应对这些挑战，我们引入了一种新的混合模型，该模型利用经验模态分解（EMD）进行信号分解和分析，有效地克服了传统方法的局限性。EMD尤其具有优势，因为它将复杂信号分解为内禀模态函数（IMFs），从而能够在长序列中更准确地检测异常。在EMD的基础上，采用高斯混合模型（GMM）精确识别各种故障模式，为异常检测奠定了坚实的基础。双向长短期记忆（BiLSTM）网络通过捕获两个方向的时间依赖性、整合关键时空信息和提高预测精度进一步增强了模型。实验结果表明，这种集成EMD-GMM-BiLSTM方法不仅在异常检测方面具有很高的灵敏度和准确性，而且比编码器-解码器模型或变压器等更复杂的框架更简单、更高效。该方法通过提供必要的安全保障，确保了直升机的运行安全，支持低空经济活动的广泛采用。

{"title":"An EMD-based forecasting framework integrating GMM and BiLSTM for helicopter engine anomaly detection","authors":"Qi Shen , Jingwei Guo , Yihui Tian , Zhen-Song Chen","doi":"10.1016/j.jii.2025.101003","DOIUrl":"10.1016/j.jii.2025.101003","url":null,"abstract":"<div><div>The safety of helicopter operations is paramount, yet early signs of potential failures often go undetected, highlighting the need for robust signal alert systems during flights. Detecting anomalies in helicopter engine behavior through vibration analysis is critically important due to the long-sequence nature and complexity of the data, which present significant challenges for real-time assessment and are not adequately addressed by traditional methods such as preset thresholds or basic statistical models, as these approaches struggle to capture intricate spatiotemporal dependencies and overlapping fault patterns in real-world scenarios. To address these challenges, we introduce a novel hybrid model that leverages Empirical Mode Decomposition (EMD) for signal decomposition and analysis, effectively overcoming the limitations of traditional approaches. EMD is particularly advantageous as it decomposes complex signals into Intrinsic Mode Functions (IMFs), enabling more accurate anomaly detection in long sequences. Following EMD, the Gaussian Mixture Model (GMM) is employed to precisely recognize various fault patterns, ensuring a robust foundation for anomaly detection. Bidirectional Long Short-Term Memory (BiLSTM) networks further enhance the model by capturing temporal dependencies in both directions, integrating critical spatiotemporal information and improving predictive accuracy. Experimental results demonstrate that this integrated EMD-GMM-BiLSTM approach is not only highly sensitive and accurate in detecting anomalies but also significantly simpler and more efficient than more complex frameworks such as encoder-decoder models or Transformers. This method ensures the operational safety of helicopters and supports the broader adoption of low-altitude economic activities by providing essential safety guarantees.</div></div>","PeriodicalId":55975,"journal":{"name":"Journal of Industrial Information Integration","volume":"48 ","pages":"Article 101003"},"PeriodicalIF":10.4,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145383754","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 Certificateless Aggregate G+G Signature Scheme with Intersection Method for Efficiency Improvement in Smart Grids 一种面向智能电网效率提升的无证书聚合G+G签名方案

IF 10.4 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Journal of Industrial Information Integration

Pub Date : 2025-11-01 DOI: 10.1016/j.jii.2025.100991

Songshou Dong , Yanqing Yao , Huaxiong Wang

Smart grids (SGs) can greatly improve the efficiency, reliability, and sustainability of traditional grids. In an industrial SG, real-time user-side metering data may be frequently collected for monitoring and controlling electricity consumption. To reduce the burden on SGs, most existing privacy-preserving schemes use aggregated signatures to ensure the integrity of the message and improve communication efficiency. In CRYPTO ’24, Marius et al. proposed an aggregating Falcon signature scheme LaBRADOR, which is a trapdoor-based lattice signature scheme. Currently, there are two types of lattice-based signature schemes: one is a trapdoor-based signature scheme, and the other is a Fiat-Shamir-based signature scheme. There is currently no particularly efficient Fiat-Shamir-based lattice-based aggregate signature scheme. Therefore, we construct an aggregate signature scheme with constant signature size without rejection sampling under the Fiat-Shamir style based on the G+G lattice signature (ASIACRYPT ’23) and the intersection method (EUROCRYPT ’11). In addition, we make our scheme certificateless to resist malicious key generation centers and the key escrow problem, and apply our scheme to SGs. Compared with other schemes, our signature scheme has a smaller aggregated signature size (any number of signatures), less signature time, and is more secure. Finally, we demonstrate that our scheme is existentially unforgeable in the context of adaptive chosen message attacks against type I and type II adversaries in the random oracle model.

智能电网可以极大地提高传统电网的效率、可靠性和可持续性。在工业SG中，可能经常收集实时用户端计量数据以监测和控制用电量。为了减轻SGs的负担，现有的大多数隐私保护方案都使用聚合签名来保证消息的完整性，提高通信效率。在CRYPTO’24中，Marius等人提出了一种聚合猎鹰签名方案LaBRADOR，这是一种基于活门的格子签名方案。目前，基于格子的签名方案主要有两种：一种是基于trapdoor的签名方案，另一种是基于fiat - shamir的签名方案。目前还没有特别高效的基于fiat - shamir的格子聚合签名方案。因此，我们基于G+G格签名（ASIACRYPT’23）和交点方法（EUROCRYPT’11），构造了一个在菲亚特-沙米尔风格下无拒绝抽样且签名大小不变的聚合签名方案。此外，为了抵御恶意密钥生成中心和密钥托管问题，我们使我们的方案无证书化，并将我们的方案应用于SGs。与其他方案相比，我们的签名方案具有签名总大小（任意数量的签名）更小、签名时间更短、安全性更高的优点。最后，我们证明了我们的方案在随机oracle模型中针对类型I和类型II对手的自适应选择消息攻击的背景下是存在不可伪造的。

{"title":"A Certificateless Aggregate G+G Signature Scheme with Intersection Method for Efficiency Improvement in Smart Grids","authors":"Songshou Dong , Yanqing Yao , Huaxiong Wang","doi":"10.1016/j.jii.2025.100991","DOIUrl":"10.1016/j.jii.2025.100991","url":null,"abstract":"<div><div>Smart grids (SGs) can greatly improve the efficiency, reliability, and sustainability of traditional grids. In an industrial SG, real-time user-side metering data may be frequently collected for monitoring and controlling electricity consumption. To reduce the burden on SGs, most existing privacy-preserving schemes use aggregated signatures to ensure the integrity of the message and improve communication efficiency. In CRYPTO ’24, Marius et al. proposed an aggregating Falcon signature scheme LaBRADOR, which is a trapdoor-based lattice signature scheme. Currently, there are two types of lattice-based signature schemes: one is a trapdoor-based signature scheme, and the other is a Fiat-Shamir-based signature scheme. There is currently no particularly efficient Fiat-Shamir-based lattice-based aggregate signature scheme. Therefore, we construct an aggregate signature scheme with constant signature size without rejection sampling under the Fiat-Shamir style based on the G+G lattice signature (ASIACRYPT ’23) and the intersection method (EUROCRYPT ’11). In addition, we make our scheme certificateless to resist malicious key generation centers and the key escrow problem, and apply our scheme to SGs. Compared with other schemes, our signature scheme has a smaller aggregated signature size (any number of signatures), less signature time, and is more secure. Finally, we demonstrate that our scheme is existentially unforgeable in the context of adaptive chosen message attacks against type I and type II adversaries in the random oracle model.</div></div>","PeriodicalId":55975,"journal":{"name":"Journal of Industrial Information Integration","volume":"48 ","pages":"Article 100991"},"PeriodicalIF":10.4,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145383758","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 multimodal data fusion-based intelligent detection method for lump coal on underground conveyor belts in smart manufacturing 基于多模态数据融合的智能制造地下传送带块煤智能检测方法

IF 10.4 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Journal of Industrial Information Integration

Pub Date : 2025-11-01 DOI: 10.1016/j.jii.2025.100997

Le Chen , Ligang Wu , Qichao Ren

To address the challenges of low detection precision for lump coal on underground coal mine conveyor belts, this study proposes an intelligent detection method based on multimodal data fusion. The method is named YOLO DKH (YOLO Dynamic Snake Attention-KANA²-High-level Screening Feature Pyramid Network). This approach specifically targets the insufficient robustness of single-modal data under dust interference and varying lighting conditions in complex underground environments. First, a Deformable Spatial Attention (DSA) mechanism is designed, utilizing strip-shaped deformable convolution kernels along the x- and y-axes for feature extraction, which achieves adaptive geometric learning and reduces computational complexity simultaneously. Second, the KANA² dual-attention mechanism is proposed by combining regional attention with the KAN Conv module. Through B-spline smoothing and dual-branch processing, computational complexity is reduced, enhancing the fusion effect of RGB-infrared multimodal features. Then, a High-frequency Spatial Feature Pyramid Network (HSFPN) was constructed by integrating high-frequency perception modules and spatial dependency perception mechanisms to enhance multi-scale feature fusion by filtering out low-frequency background interference and capturing pixel-level spatial relationships. Finally, a comprehensive multi-modal RGB-infrared dataset comprising 9250 annotated images and 14,840 bounding boxes was constructed to provide a standardized benchmark for the development and validation of lump coal detection algorithms. The experimental results showed that the YOLO DKH model achieved 79.1 %, 74.3 %, and 77.2 % precision, recall, and [email protected], respectively, representing improvements of 6.03 %, 7.06 %, and 5.18 % compared to the baseline YOLOv11n model, while reducing the number of parameters by 2.71 %. and a 25.9 % reduction in single-image processing time to 6.1 milliseconds, providing an efficient and reliable technical solution for lump coal monitoring on underground conveyor belts in intelligent manufacturing.

针对煤矿井下传送带块煤检测精度低的问题，提出了一种基于多模态数据融合的块煤智能检测方法。该方法被命名为YOLO DKH （YOLO Dynamic Snake Attention-KANA2-High-level Screening Feature Pyramid Network）。该方法专门针对复杂地下环境中单模态数据在粉尘干扰和光照条件变化下鲁棒性不足的问题。首先，设计了一种可变形空间注意（DSA）机制，利用沿x轴和y轴的条形可变形卷积核进行特征提取，实现了自适应几何学习，同时降低了计算复杂度；其次，将区域注意与KAN转换模块相结合，提出KANA2双注意机制。通过b样条平滑和双分支处理，降低了计算复杂度，增强了红外多模态特征的融合效果。然后，结合高频感知模块和空间依赖感知机制构建高频空间特征金字塔网络（HSFPN），通过滤除低频背景干扰和捕获像素级空间关系增强多尺度特征融合；最后，构建了包含9250张带注释图像和14840个边界框的综合多模态rgb -红外数据集，为块煤检测算法的开发和验证提供了标准化基准。实验结果表明，YOLO DKH模型的准确率、召回率和[email protected]分别达到79.1%、74.3%和77.2%，与基线YOLOv11n模型相比分别提高了6.03%、7.06%和5.18%，同时减少了2.71%的参数数量。单幅图像处理时间缩短25.9%，达到6.1毫秒，为智能制造中块煤井下传送带监测提供了高效可靠的技术解决方案。

{"title":"A multimodal data fusion-based intelligent detection method for lump coal on underground conveyor belts in smart manufacturing","authors":"Le Chen , Ligang Wu , Qichao Ren","doi":"10.1016/j.jii.2025.100997","DOIUrl":"10.1016/j.jii.2025.100997","url":null,"abstract":"<div><div>To address the challenges of low detection precision for lump coal on underground coal mine conveyor belts, this study proposes an intelligent detection method based on multimodal data fusion. The method is named YOLO DKH (YOLO Dynamic Snake Attention-KANA<sup>2</sup>-High-level Screening Feature Pyramid Network). This approach specifically targets the insufficient robustness of single-modal data under dust interference and varying lighting conditions in complex underground environments. First, a Deformable Spatial Attention (DSA) mechanism is designed, utilizing strip-shaped deformable convolution kernels along the x- and y-axes for feature extraction, which achieves adaptive geometric learning and reduces computational complexity simultaneously. Second, the KANA<sup>2</sup> dual-attention mechanism is proposed by combining regional attention with the KAN Conv module. Through B-spline smoothing and dual-branch processing, computational complexity is reduced, enhancing the fusion effect of RGB-infrared multimodal features. Then, a High-frequency Spatial Feature Pyramid Network (HSFPN) was constructed by integrating high-frequency perception modules and spatial dependency perception mechanisms to enhance multi-scale feature fusion by filtering out low-frequency background interference and capturing pixel-level spatial relationships. Finally, a comprehensive multi-modal RGB-infrared dataset comprising 9250 annotated images and 14,840 bounding boxes was constructed to provide a standardized benchmark for the development and validation of lump coal detection algorithms. The experimental results showed that the YOLO DKH model achieved 79.1 %, 74.3 %, and 77.2 % precision, recall, and [email protected], respectively, representing improvements of 6.03 %, 7.06 %, and 5.18 % compared to the baseline YOLOv11n model, while reducing the number of parameters by 2.71 %. and a 25.9 % reduction in single-image processing time to 6.1 milliseconds, providing an efficient and reliable technical solution for lump coal monitoring on underground conveyor belts in intelligent manufacturing.</div></div>","PeriodicalId":55975,"journal":{"name":"Journal of Industrial Information Integration","volume":"48 ","pages":"Article 100997"},"PeriodicalIF":10.4,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145383455","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

Special issue on “Industrial information integration in space informatics” 《空间信息学中的产业信息集成》特刊

IF 10.4 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Journal of Industrial Information Integration

Pub Date : 2025-11-01 DOI: 10.1016/j.jii.2025.100957

Yuk Ming Tang , Andrew W.H. Ip , Kai Leung Yung , Zhuming Bi , Zhili Sun

引用次数: 0

A literature review and bibliometric analysis of 50 years of optimization approaches applied to the order batching problem 50年来优化方法应用于有序批处理问题的文献回顾与计量分析

IF 10.4 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Journal of Industrial Information Integration

Pub Date : 2025-11-01 DOI: 10.1016/j.jii.2025.100993

Anderson Rogério Faia Pinto , Esra Boz , Rafael Henrique Faia Pinto , Marcelo Seido Nagano

This manuscript presents a literature review with a bibliometric analysis on the Order Batching Problem (OBP). The research analyzed Literature Reviews (30) and Picking Optimization Methods (138). Most approaches focus on hypothetical warehouses with static (offline) orders and are configured as the classical OBP. These warehouses feature rectangular layouts (single-block and parallel aisles) with low-level picker-to-parts systems and one Pick-up and Drop-off. Most effective solutions have emerged from metaheuristics in conjunction with constructive heuristics, and the most frequently utilized techniques are the Genetic Algorithm and Variable Neighborhood Search. The main performance indicators are the Total Picking Time, the Total Routing Distance, and the Computational Processing Time. The bibliometric analyses encompassed Journals (77), Universities (169), and Researchers (331). Most publications originate from journals in Europe and North America. The countries with the highest concentration of universities and researchers are the United States and China. Nevertheless, authorship analysis shows that China and Germany outperform the United States. The continents with the largest number of researchers are Asia and Europe. However, a ranking by authorship reveals that the researchers with the most publications are from Europe and South America. This manuscript presents the state of the art, demonstrates advancements in the field, identifies research interests, examines customer service level requirements and warehouse efficiency, and addresses the gap for more comprehensive bibliometric analyses on OBP. Formulating Picking Optimization Methods better adapted and capable of addressing real-world trade-offs constitutes the primary challenge and the most promising future approaches for the OBP.

这篇手稿提出了一个文献综述与文献计量分析的顺序批问题（OBP）。研究分析了文献综述（30）和选择优化方法（138）。大多数方法关注具有静态（离线）订单的假想仓库，并将其配置为经典OBP。这些仓库的特点是矩形布局（单块平行通道），具有低级的拣货到零件系统和一个取货和落货系统。最有效的解决方案是从元启发式与建设性启发式结合出现的，最常用的技术是遗传算法和变量邻域搜索。主要性能指标有总拣货时间、总路由距离和计算处理时间。文献计量分析包括期刊（77）、大学（169）和研究人员（331）。大多数出版物来自欧洲和北美的期刊。大学和研究人员最集中的国家是美国和中国。然而，作者分析显示，中国和德国的表现优于美国。拥有最多研究人员的大陆是亚洲和欧洲。然而，根据作者排名显示，发表论文最多的研究人员来自欧洲和南美。这份手稿介绍了最新的技术，展示了该领域的进步，确定了研究兴趣，检查了客户服务水平要求和仓库效率，并解决了OBP上更全面的文献计量分析的差距。制定更好地适应和能够解决现实权衡的采油优化方法是OBP面临的主要挑战，也是未来最有希望的方法。

{"title":"A literature review and bibliometric analysis of 50 years of optimization approaches applied to the order batching problem","authors":"Anderson Rogério Faia Pinto , Esra Boz , Rafael Henrique Faia Pinto , Marcelo Seido Nagano","doi":"10.1016/j.jii.2025.100993","DOIUrl":"10.1016/j.jii.2025.100993","url":null,"abstract":"<div><div>This manuscript presents a literature review with a bibliometric analysis on the Order Batching Problem (OBP). The research analyzed Literature Reviews (30) and Picking Optimization Methods (138). Most approaches focus on hypothetical warehouses with static (offline) orders and are configured as the classical OBP. These warehouses feature rectangular layouts (single-block and parallel aisles) with low-level picker-to-parts systems and one Pick-up and Drop-off. Most effective solutions have emerged from metaheuristics in conjunction with constructive heuristics, and the most frequently utilized techniques are the Genetic Algorithm and Variable Neighborhood Search. The main performance indicators are the Total Picking Time, the Total Routing Distance, and the Computational Processing Time. The bibliometric analyses encompassed Journals (77), Universities (169), and Researchers (331). Most publications originate from journals in Europe and North America. The countries with the highest concentration of universities and researchers are the United States and China. Nevertheless, authorship analysis shows that China and Germany outperform the United States. The continents with the largest number of researchers are Asia and Europe. However, a ranking by authorship reveals that the researchers with the most publications are from Europe and South America. This manuscript presents the state of the art, demonstrates advancements in the field, identifies research interests, examines customer service level requirements and warehouse efficiency, and addresses the gap for more comprehensive bibliometric analyses on OBP. Formulating Picking Optimization Methods better adapted and capable of addressing real-world trade-offs constitutes the primary challenge and the most promising future approaches for the OBP.</div></div>","PeriodicalId":55975,"journal":{"name":"Journal of Industrial Information Integration","volume":"48 ","pages":"Article 100993"},"PeriodicalIF":10.4,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145382968","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

F2P-Net: A Hybrid Prompt-Enhanced Dual-Branch Cooperative Network for Industrial Surface Defect Segmentation with Limited Data F2P-Net：用于有限数据下工业表面缺陷分割的混合快速增强双分支协作网络

IF 10.4 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Journal of Industrial Information Integration

Pub Date : 2025-11-01 DOI: 10.1016/j.jii.2025.100986

Kerong Yan , Shuai Chen , Min Xu , Peiye Sun , Rui Wang

Industrial surface defect detection is constrained by the scarcity of defective samples and by the insufficient capacity of current segmentation methods to precisely delineate defect boundaries. To address these challenges, we propose F2P-Net, a few-sample, highly precise industrial surface defect segmentation framework composed of three core modules. ViCNet (ViT and CNN collaborative encoder network) integrates a vision transformer backbone with an auxiliary convolutional branch to retain robust large-model priors while enhancing sensitivity to fine-scale textures and local irregularities. AFDec (automated geometric prompt and multi-scale feature fusion decoder) employs automated geometric prompts to localize potential defect regions and fuses hierarchical multi-scale features to improve boundary delineation and mask consistency. EVPT (edge-enhanced visual prompt tuning) is a fine-tuning module incorporating edge-explicit visual prompt to facilitate effective industrial domain adaptation of large vision models. The proposed method achieves considerable performance over existing full-data training approaches in metrics including mAP, Recall, and IoU using only 1.76 %∼3.06 % of training images across NEU_Seg, MT, KolektorSDD2, and DAGM2007 datasets. Under full-data training, it attains state-of-the-art segmentation accuracies with IoU scores of 86.03 %, 92.57 %, 78.77 %, and 82.55 %, respectively. The network provides a novel solution for industrial applications with few-sample, high-precision defect segmentation. Code is available at https://github.com/kerongYan/F2P-Net.

工业表面缺陷检测受到缺陷样本稀缺和当前分割方法精确描绘缺陷边界能力不足的限制。为了解决这些挑战，我们提出了F2P-Net，这是一个由三个核心模块组成的少数样本，高精度工业表面缺陷分割框架。ViCNet （ViT和CNN协同编码器网络）集成了视觉变换主干和辅助卷积分支，在保留鲁棒大模型先验的同时增强了对精细尺度纹理和局部不规则性的敏感性。AFDec （automated geometric prompt and multi-scale feature fusion decoder）采用自动几何提示定位潜在缺陷区域，并融合分层多尺度特征，提高边界划定和掩码一致性。EVPT（边缘增强视觉提示调整）是一种包含边缘显式视觉提示的微调模块，用于促进大型视觉模型的有效工业领域适应。该方法仅使用NEU_Seg、MT、KolektorSDD2和DAGM2007数据集上1.76% ~ 3.06%的训练图像，在mAP、Recall和IoU等指标上比现有的全数据训练方法取得了相当大的性能。在全数据训练下，IoU分数分别为86.03%、92.57%、78.77%和82.55%，达到了最先进的分割准确率。该网络为工业应用中少样本、高精度的缺陷分割提供了一种新的解决方案。代码可从https://github.com/kerongYan/F2P-Net获得。

{"title":"F2P-Net: A Hybrid Prompt-Enhanced Dual-Branch Cooperative Network for Industrial Surface Defect Segmentation with Limited Data","authors":"Kerong Yan , Shuai Chen , Min Xu , Peiye Sun , Rui Wang","doi":"10.1016/j.jii.2025.100986","DOIUrl":"10.1016/j.jii.2025.100986","url":null,"abstract":"<div><div>Industrial surface defect detection is constrained by the scarcity of defective samples and by the insufficient capacity of current segmentation methods to precisely delineate defect boundaries. To address these challenges, we propose <strong>F2P-Net</strong>, a few-sample, highly precise industrial surface defect segmentation framework composed of three core modules. <strong>ViCNet</strong> (ViT and CNN collaborative encoder network) integrates a vision transformer backbone with an auxiliary convolutional branch to retain robust large-model priors while enhancing sensitivity to fine-scale textures and local irregularities. <strong>AFDec</strong> (automated geometric prompt and multi-scale feature fusion decoder) employs automated geometric prompts to localize potential defect regions and fuses hierarchical multi-scale features to improve boundary delineation and mask consistency. <strong>EVPT</strong> (edge-enhanced visual prompt tuning) is a fine-tuning module incorporating edge-explicit visual prompt to facilitate effective industrial domain adaptation of large vision models. The proposed method achieves considerable performance over existing full-data training approaches in metrics including mAP, Recall, and IoU using only 1.76 %∼3.06 % of training images across NEU_Seg, MT, KolektorSDD2, and DAGM2007 datasets. Under full-data training, it attains state-of-the-art segmentation accuracies with IoU scores of 86.03 %, 92.57 %, 78.77 %, and 82.55 %, respectively. The network provides a novel solution for industrial applications with few-sample, high-precision defect segmentation. Code is available at <span><span>https://github.com/kerongYan/F2P-Net</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":55975,"journal":{"name":"Journal of Industrial Information Integration","volume":"48 ","pages":"Article 100986"},"PeriodicalIF":10.4,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145416549","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

Residual group attention network with depthwise separable convolutional neural network for credit evaluation and early warning in finance 基于深度可分卷积神经网络的残差群注意网络在金融信用评估与预警中的应用

IF 10.4 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Journal of Industrial Information Integration

Pub Date : 2025-11-01 DOI: 10.1016/j.jii.2025.101002

Jayaraman Kumarappan , Ammupriya A , S. Vijay Mallikraj , Mohammed Al-Farouni

The integration of big data technology with the Internet of Things (IoT) in finance presents an opportunity to enhance credit evaluation and early warning systems for credit risks. Traditional methods face challenges in processing multi-source heterogeneous financial data, leading to inaccuracies in credit risk assessments. This paper proposes an advanced approach using Local and Global Depth Normalization for data preprocessing, which enhances data quality and consistency. For feature extraction, a Spike-Driven Transformer effectively captures intricate patterns in financial transactions. A Residual Group Attention Network with Depthwise Separable Convolutional Neural Network (RGA-DSCNN) is then employed for classification, providing high accuracy in credit risk assessment by capturing both local and global feature dependencies. To further enhance the model's performance, the Mountaineering Team-Based Optimization technique is applied to optimize the parameters of the RGA-DSCNN. The proposed model is evaluated using IoT financial data consisting of 26 indicators, and factor analysis is conducted using SPSS26.0 software for initial validation. The results demonstrate that this method significantly outperforms existing techniques, achieving more precise credit risk assessments and reducing the probability of Type I and Type II errors in credit evaluation. This approach offers an effective and robust solution for banks, facilitating more accurate credit risk management, reducing potential losses, and improving profitability in IoT finance operations.

大数据技术与金融物联网（IoT）的融合为加强信用评估和信用风险预警系统提供了机会。传统方法在处理多源异构金融数据时面临挑战，导致信用风险评估不准确。提出了一种采用局部深度归一化和全局深度归一化进行数据预处理的方法，提高了数据的质量和一致性。对于特征提取，Spike-Driven Transformer可以有效地捕获金融交易中的复杂模式。然后采用深度可分离卷积神经网络（RGA-DSCNN）的残差群注意网络进行分类，通过捕获局部和全局特征依赖关系来提高信用风险评估的准确性。为了进一步提高模型的性能，采用登山队优化技术对RGA-DSCNN的参数进行优化。采用包含26个指标的物联网财务数据对模型进行评价，并采用SPSS26.0软件进行因子分析进行初步验证。结果表明，该方法显著优于现有技术，实现了更精确的信用风险评估，并降低了信用评估中I类和II类错误的概率。这种方法为银行提供了一种有效而稳健的解决方案，有助于更准确地管理信贷风险，减少潜在损失，提高物联网金融运营的盈利能力。

{"title":"Residual group attention network with depthwise separable convolutional neural network for credit evaluation and early warning in finance","authors":"Jayaraman Kumarappan , Ammupriya A , S. Vijay Mallikraj , Mohammed Al-Farouni","doi":"10.1016/j.jii.2025.101002","DOIUrl":"10.1016/j.jii.2025.101002","url":null,"abstract":"<div><div>The integration of big data technology with the Internet of Things (IoT) in finance presents an opportunity to enhance credit evaluation and early warning systems for credit risks. Traditional methods face challenges in processing multi-source heterogeneous financial data, leading to inaccuracies in credit risk assessments. This paper proposes an advanced approach using Local and Global Depth Normalization for data preprocessing, which enhances data quality and consistency. For feature extraction, a Spike-Driven Transformer effectively captures intricate patterns in financial transactions. A Residual Group Attention Network with Depthwise Separable Convolutional Neural Network (RGA-DSCNN) is then employed for classification, providing high accuracy in credit risk assessment by capturing both local and global feature dependencies. To further enhance the model's performance, the Mountaineering Team-Based Optimization technique is applied to optimize the parameters of the RGA-DSCNN. The proposed model is evaluated using IoT financial data consisting of 26 indicators, and factor analysis is conducted using SPSS26.0 software for initial validation. The results demonstrate that this method significantly outperforms existing techniques, achieving more precise credit risk assessments and reducing the probability of Type I and Type II errors in credit evaluation. This approach offers an effective and robust solution for banks, facilitating more accurate credit risk management, reducing potential losses, and improving profitability in IoT finance operations.</div></div>","PeriodicalId":55975,"journal":{"name":"Journal of Industrial Information Integration","volume":"48 ","pages":"Article 101002"},"PeriodicalIF":10.4,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145383755","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

Sustainable building material supplier assessment in Pythagorean neutrosophic setting using ITARA and MACONT methods 使用ITARA和MACONT方法对毕达哥拉斯中性环境下的可持续建筑材料供应商进行评估

IF 10.4 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Journal of Industrial Information Integration

Pub Date : 2025-11-01 DOI: 10.1016/j.jii.2025.101000

Tapas Kumar Paul , Madhumangal Pal

Supplier selection plays a pivotal role in the construction industry, particularly for companies that seek to harmonize sustainability, innovation, and quality in their operations. This paper introduces a novel multi-criteria decision-making (MCDM) framework that combines the Indifference Threshold-Based Attribute Ratio Analysis (ITARA) and the Mixed Aggregation by Comprehensive Normalization Technique (MACONT) within a Pythagorean neutrosophic environment. The model is designed to assess suppliers using four critical dimensions: product performance, innovation in the supply chain, service effectiveness, and environmental sustainability, tailored to meet the strategic needs of a prominent Indian company. By utilizing Pythagorean neutrosophic sets, the proposed approach effectively addresses uncertainties and ambiguities in expert evaluations, resulting in a more adaptable and dependable decision-making process. The application of this hybrid ITARA-MACONT method shows its effectiveness in selecting top-performing suppliers who align with these multifaceted requirements, thereby enabling companies to make decisions that prioritize both sustainability and innovation. The practical application of this method illustrates the model’s feasibility and robustness. Moreover, the proposed framework supports industrial information integration by enabling the structured fusion of heterogeneous expert judgments and quantitative performance metrics, critical for informed decision-making in digital supply chains.

供应商选择在建筑行业中起着关键作用，特别是对于那些寻求在运营中协调可持续性、创新和质量的公司。提出了一种基于无差异阈值的属性比分析（ITARA）和综合归一化混合聚合技术（MACONT）的多准则决策框架。该模型旨在通过四个关键维度来评估供应商：产品性能、供应链创新、服务效率和环境可持续性，以满足一家知名印度公司的战略需求。通过利用毕达哥拉斯中性集，该方法有效地解决了专家评估中的不确定性和模糊性，从而使决策过程更具适应性和可靠性。这种ITARA-MACONT混合方法的应用显示了其在选择符合这些多方面要求的最佳供应商方面的有效性，从而使公司能够做出优先考虑可持续性和创新的决策。该方法的实际应用证明了该模型的可行性和鲁棒性。此外，该框架通过实现异构专家判断和定量绩效指标的结构化融合来支持工业信息集成，这对数字供应链中的知情决策至关重要。

{"title":"Sustainable building material supplier assessment in Pythagorean neutrosophic setting using ITARA and MACONT methods","authors":"Tapas Kumar Paul , Madhumangal Pal","doi":"10.1016/j.jii.2025.101000","DOIUrl":"10.1016/j.jii.2025.101000","url":null,"abstract":"<div><div>Supplier selection plays a pivotal role in the construction industry, particularly for companies that seek to harmonize sustainability, innovation, and quality in their operations. This paper introduces a novel multi-criteria decision-making (MCDM) framework that combines the Indifference Threshold-Based Attribute Ratio Analysis (ITARA) and the Mixed Aggregation by Comprehensive Normalization Technique (MACONT) within a Pythagorean neutrosophic environment. The model is designed to assess suppliers using four critical dimensions: product performance, innovation in the supply chain, service effectiveness, and environmental sustainability, tailored to meet the strategic needs of a prominent Indian company. By utilizing Pythagorean neutrosophic sets, the proposed approach effectively addresses uncertainties and ambiguities in expert evaluations, resulting in a more adaptable and dependable decision-making process. The application of this hybrid ITARA-MACONT method shows its effectiveness in selecting top-performing suppliers who align with these multifaceted requirements, thereby enabling companies to make decisions that prioritize both sustainability and innovation. The practical application of this method illustrates the model’s feasibility and robustness. Moreover, the proposed framework supports industrial information integration by enabling the structured fusion of heterogeneous expert judgments and quantitative performance metrics, critical for informed decision-making in digital supply chains.</div></div>","PeriodicalId":55975,"journal":{"name":"Journal of Industrial Information Integration","volume":"48 ","pages":"Article 101000"},"PeriodicalIF":10.4,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145442110","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

Task-level failure diagnosis in process control systems under cyberattack based on multilevel business process models 基于多级业务流程模型的网络攻击下过程控制系统任务级故障诊断

IF 10.4 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Journal of Industrial Information Integration

Pub Date : 2025-11-01 DOI: 10.1016/j.jii.2025.101001

Xuqing Liang , Chunjie Zhou , Yu-Chu Tian , Peihang Xu , Minglu Wang , Xin Du

Ensuring the secure operation of process control systems (PCSs) is critical in the face of increasing cyberattacks. Diagnosing the causes of safety protection failures and predicting potential failure paths during cyberattacks can help maintain the system stability and security. However, cyberattacks often target not only assets but also control tasks or disrupt business processes, rendering existing asset-based failure diagnosis methods insufficient. To address this gap, this paper presents a novel task-level failure diagnosis approach based on multilevel business process models (BPMs) for PCSs. The model is constructed based on Business Process Model and Notation (BPMN) 2.0 by analyzing the interactions between devices, tasks, and business processes. It is then mapped to a multilevel Bayesian network (BN) for quantitative analysis under cyberattacks. Finally, the proposed approach is validated through a simulated distillation unit. Results show that cause diagnosis confirms sensing task failure (40.7%) as the main cause of business process failures, primarily due to data transmission failure (54.5%). The method effectively diagnoses failure causes at the task level and predicting potential system safety protection failure paths.

面对日益增多的网络攻击，确保过程控制系统（pcs）的安全运行至关重要。在网络攻击中，诊断安全防护故障的原因，预测潜在的故障路径，有助于维护系统的稳定性和安全性。然而，网络攻击往往不仅针对资产，还针对控制任务或破坏业务流程，使得现有的基于资产的故障诊断方法不足。为了解决这一问题，本文提出了一种基于多层次业务流程模型（bpm）的任务级故障诊断方法。通过分析设备、任务和业务流程之间的交互，该模型基于业务流程模型和符号（BPMN） 2.0构建。然后将其映射到多层次贝叶斯网络（BN）中进行网络攻击下的定量分析。最后，通过模拟蒸馏装置对该方法进行了验证。结果表明，原因诊断确认感知任务失败（40.7%）是业务流程失败的主要原因，主要原因是数据传输失败（54.5%）。该方法能有效地在任务级诊断故障原因，预测系统安全保护的潜在故障路径。

{"title":"Task-level failure diagnosis in process control systems under cyberattack based on multilevel business process models","authors":"Xuqing Liang , Chunjie Zhou , Yu-Chu Tian , Peihang Xu , Minglu Wang , Xin Du","doi":"10.1016/j.jii.2025.101001","DOIUrl":"10.1016/j.jii.2025.101001","url":null,"abstract":"<div><div>Ensuring the secure operation of process control systems (PCSs) is critical in the face of increasing cyberattacks. Diagnosing the causes of safety protection failures and predicting potential failure paths during cyberattacks can help maintain the system stability and security. However, cyberattacks often target not only assets but also control tasks or disrupt business processes, rendering existing asset-based failure diagnosis methods insufficient. To address this gap, this paper presents a novel task-level failure diagnosis approach based on multilevel business process models (BPMs) for PCSs. The model is constructed based on Business Process Model and Notation (BPMN) 2.0 by analyzing the interactions between devices, tasks, and business processes. It is then mapped to a multilevel Bayesian network (BN) for quantitative analysis under cyberattacks. Finally, the proposed approach is validated through a simulated distillation unit. Results show that cause diagnosis confirms sensing task failure (40.7%) as the main cause of business process failures, primarily due to data transmission failure (54.5%). The method effectively diagnoses failure causes at the task level and predicting potential system safety protection failure paths.</div></div>","PeriodicalId":55975,"journal":{"name":"Journal of Industrial Information Integration","volume":"48 ","pages":"Article 101001"},"PeriodicalIF":10.4,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145412200","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