Journal of Industrial Information Integration最新文献_第5页

A distributed extended reality escape method for layered underground infrastructure based on AI game engine 基于AI游戏引擎的分层地下基础设施分布式扩展现实逃生方法

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

Journal of Industrial Information Integration

Pub Date : 2025-11-16 DOI: 10.1016/j.jii.2025.101015

Wei Li , Linbing Wang , Maogui Sun , Dengcai Yin , Yajian Wang , Xiang Zhou , Yongming Wang , Zhoujing Ye

As the structural carrier of mineral resources, underground mine is a typical artificial large layered underground infrastructure. The safety of mining systems remains a critical concern for nations worldwide. Based on the environmental characteristics of underground mines, the accompanying safety issues are evident. Conventional personnel evacuation drills for mine disasters often fail to create effective disaster evolution memories for people. When a real accident occurs, people cannot escape efficiently in a panic state, which reduces survival probability. To solve this problem, an escape space connection algorithm is developed based on the physical information and management rules in this study, and it is used to drive the extended reality escape system by the game engine. Firstly, this study takes the water-inrush accidents of underground layered mines as the engineering research object and background, the characteristics of water-inrush accidents evolution and personnel evacuation are systematically analyzed based on the scenario construction theory. Secondly, this study develops an escape space connection algorithm by integrating the two-dimensional A* algorithm and the connection weights of escape spaces based on the spatial geometric information and escape strategy of layered mines. Thirdly, a distributed extended reality (XR) human-computer interaction system is developed for escape path guidance in real environments based on the spatial structure characteristics of layered mines and the escape space connection algorithm. Finally, application testing is conducted in the experimental mine to analyze the system performance and future application potential. This study provides a comprehensive technical framework for personnel evacuation in layered underground infrastructure during evolutionary accidents, and the theories and systems involved are universal. In addition, this method can be used as a new, low-cost and efficient digital reference system for personnel safety emergency drills in underground infrastructure.

地下矿山作为矿产资源的结构载体，是典型的人工大型层状地下基础设施。采矿系统的安全仍然是世界各国关切的一个重大问题。根据地下矿山的环境特点，伴随而来的安全问题是显而易见的。传统的矿难人员疏散演练往往不能为人们创造有效的灾害演化记忆。当真正发生事故时，人们在恐慌状态下无法有效逃生，降低了生存概率。为解决这一问题，本研究开发了一种基于物理信息和管理规则的逃生空间连接算法，并通过游戏引擎驱动扩展现实逃生系统。首先，本研究以地下分层矿山突水事故为工程研究对象和背景，基于场景构建理论系统分析了突水事故演化和人员疏散的特征。其次，基于层状矿山的空间几何信息和逃生策略，将二维A*算法与逃生空间的连接权值相结合，开发了逃生空间连接算法。第三，基于层状矿井空间结构特点和逃生空间连接算法，开发了面向真实环境的分布式扩展现实（XR）逃生路径引导人机交互系统。最后在实验矿山进行了应用测试，分析了系统的性能和未来的应用潜力。本研究为演化事故中分层地下基础设施人员疏散提供了一个全面的技术框架，所涉及的理论和系统具有普遍性。该方法可作为地下基础设施人员安全应急演练的一种新型、低成本、高效的数字参考系统。

{"title":"A distributed extended reality escape method for layered underground infrastructure based on AI game engine","authors":"Wei Li , Linbing Wang , Maogui Sun , Dengcai Yin , Yajian Wang , Xiang Zhou , Yongming Wang , Zhoujing Ye","doi":"10.1016/j.jii.2025.101015","DOIUrl":"10.1016/j.jii.2025.101015","url":null,"abstract":"<div><div>As the structural carrier of mineral resources, underground mine is a typical artificial large layered underground infrastructure. The safety of mining systems remains a critical concern for nations worldwide. Based on the environmental characteristics of underground mines, the accompanying safety issues are evident. Conventional personnel evacuation drills for mine disasters often fail to create effective disaster evolution memories for people. When a real accident occurs, people cannot escape efficiently in a panic state, which reduces survival probability. To solve this problem, an escape space connection algorithm is developed based on the physical information and management rules in this study, and it is used to drive the extended reality escape system by the game engine. Firstly, this study takes the water-inrush accidents of underground layered mines as the engineering research object and background, the characteristics of water-inrush accidents evolution and personnel evacuation are systematically analyzed based on the scenario construction theory. Secondly, this study develops an escape space connection algorithm by integrating the two-dimensional A* algorithm and the connection weights of escape spaces based on the spatial geometric information and escape strategy of layered mines. Thirdly, a distributed extended reality (XR) human-computer interaction system is developed for escape path guidance in real environments based on the spatial structure characteristics of layered mines and the escape space connection algorithm. Finally, application testing is conducted in the experimental mine to analyze the system performance and future application potential. This study provides a comprehensive technical framework for personnel evacuation in layered underground infrastructure during evolutionary accidents, and the theories and systems involved are universal. In addition, this method can be used as a new, low-cost and efficient digital reference system for personnel safety emergency drills in underground infrastructure.</div></div>","PeriodicalId":55975,"journal":{"name":"Journal of Industrial Information Integration","volume":"49 ","pages":"Article 101015"},"PeriodicalIF":10.4,"publicationDate":"2025-11-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145531166","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

Category-controllable and high-fidelity 3D defect synthesis for Embodied Intelligence-based industrial inspection 面向具体智能工业检测的类别可控高保真三维缺陷综合

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

Journal of Industrial Information Integration

Pub Date : 2025-11-15 DOI: 10.1016/j.jii.2025.101016

Ting Li , Di Li , Chunhua Zhang , Peng Chi , Ziren Luo

Embodied Intelligence (EI) integrates perception, cognition, and action within manufacturing systems, enabling on-device learning and human-machine collaboration. For surface defect inspection, this requires real-time reasoning over subtle 3D geometries and continuous self-improvement using high-quality training data. However, current point cloud generation methods fall short in synthesizing 3D defects due to inefficient single-class generation, lack of pixel-level annotations, and poor diversity. We propose Category-Controllable and High-Fidelity Generative Adversarial Network (CFGAN) to address these issues. CFGAN generates paired RGB and depth defect images with controllable categories and pure backgrounds, enabling multi-class synthesis and facilitating pixel-level annotation. A gradient-adaptive Poisson fusion method ensures seamless blending of generated RGB and depth defects into normal backgrounds, while domain transfer and depth mapping modules are further applied to preserve the consistency and reliability of the generated depth. Moreover, by sampling random latent codes, CFGAN produces diverse defect samples. Finally, spatial alignment of defect images maps 2D features into 3D space, resulting in realistic defect point clouds. The effectiveness of the proposed method is validated through experiments on fruit, metal, and plastic objects. In addition, our framework enables zero-shot inspection by transferring defects across datasets with different backgrounds but similar defects, achieving an Overall Accuracy of 0.9736. Our work provides diverse, well-annotated point cloud defects, enhancing the adaptability and autonomy of EI inspection systems.

具身智能（EI）集成了感知、认知和制造系统中的行动，实现了设备上的学习和人机协作。对于表面缺陷检测，这需要对细微的3D几何图形进行实时推理，并使用高质量的训练数据进行持续的自我改进。然而，目前的点云生成方法由于单类生成效率低、缺乏像素级注释、多样性差等原因，在合成三维缺陷方面存在不足。我们提出了类别可控和高保真生成对抗网络（CFGAN）来解决这些问题。CFGAN生成具有可控类别和纯背景的RGB和深度缺陷配对图像，实现多类别合成，便于像素级标注。采用梯度自适应泊松融合方法将生成的RGB缺陷和深度缺陷无缝融合到正常背景中，并进一步应用域转移和深度映射模块来保持生成深度的一致性和可靠性。此外，通过对随机潜在码进行采样，CFGAN产生了多种缺陷样本。最后，对缺陷图像进行空间对齐，将二维特征映射到三维空间，得到逼真的缺陷点云。通过水果、金属和塑料物体的实验验证了该方法的有效性。此外，我们的框架通过在具有不同背景但相似缺陷的数据集之间转移缺陷来实现零射击检查，实现了0.9736的总体精度。我们的工作提供了多样化、良好注释的点云缺陷，增强了EI检测系统的适应性和自主性。

{"title":"Category-controllable and high-fidelity 3D defect synthesis for Embodied Intelligence-based industrial inspection","authors":"Ting Li , Di Li , Chunhua Zhang , Peng Chi , Ziren Luo","doi":"10.1016/j.jii.2025.101016","DOIUrl":"10.1016/j.jii.2025.101016","url":null,"abstract":"<div><div>Embodied Intelligence (EI) integrates perception, cognition, and action within manufacturing systems, enabling on-device learning and human-machine collaboration. For surface defect inspection, this requires real-time reasoning over subtle 3D geometries and continuous self-improvement using high-quality training data. However, current point cloud generation methods fall short in synthesizing 3D defects due to inefficient single-class generation, lack of pixel-level annotations, and poor diversity. We propose Category-Controllable and High-Fidelity Generative Adversarial Network (CFGAN) to address these issues. CFGAN generates paired RGB and depth defect images with controllable categories and pure backgrounds, enabling multi-class synthesis and facilitating pixel-level annotation. A gradient-adaptive Poisson fusion method ensures seamless blending of generated RGB and depth defects into normal backgrounds, while domain transfer and depth mapping modules are further applied to preserve the consistency and reliability of the generated depth. Moreover, by sampling random latent codes, CFGAN produces diverse defect samples. Finally, spatial alignment of defect images maps 2D features into 3D space, resulting in realistic defect point clouds. The effectiveness of the proposed method is validated through experiments on fruit, metal, and plastic objects. In addition, our framework enables zero-shot inspection by transferring defects across datasets with different backgrounds but similar defects, achieving an Overall Accuracy of 0.9736. Our work provides diverse, well-annotated point cloud defects, enhancing the adaptability and autonomy of EI inspection systems.</div></div>","PeriodicalId":55975,"journal":{"name":"Journal of Industrial Information Integration","volume":"49 ","pages":"Article 101016"},"PeriodicalIF":10.4,"publicationDate":"2025-11-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145531167","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

Physics-informed continuous-time reinforcement learning with data-driven approach for robotic arm manipulation 基于物理信息的连续时间强化学习与数据驱动的机械臂操作方法

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

Journal of Industrial Information Integration

Pub Date : 2025-11-11 DOI: 10.1016/j.jii.2025.101008

Jin-Qiang Wang , Lirong Song , Jun Shen , Binbin Yong , Xiaoteng Han , Yuanbo Jiang , Mona Raoufi , Qingguo Zhou

Deep reinforcement learning (DRL) plays a crucial role in complex sequential decision-making tasks. However, existing data-driven DRL methods primarily rely on an empirical risk minimization (ERM) strategy to fit optimal value function models. This approach often neglects the environment’s dynamical system properties, which in turn leads to an inadequate consideration of the structural risk minimization (SRM) strategy. To address this limitation, this paper proposes a physics-informed continuous-time reinforcement learning (PICRL) to validate model effectiveness from both ERM and SRM perspectives. Specifically, we begin by theoretically analyzing the mechanism of SRM in reinforcement learning models. Then, physics information is integrated into both discrete and continuous reinforcement learning algorithms for comparative experiments. Finally, we systematically examine the effects of various physics-informed and boundary constraints on these two learning frameworks. Experimental results on the PandaGym demonstrate that the proposed method achieves comparable or superior performance in both discrete and continuous-time reinforcement learning frameworks. This provides strong evidence for its significant advantages in learning control policies for dynamical systems with small time intervals.

深度强化学习（DRL）在复杂的序列决策任务中起着至关重要的作用。然而，现有的数据驱动DRL方法主要依靠经验风险最小化（ERM）策略来拟合最优价值函数模型。这种方法往往忽略了环境的动力系统特性，从而导致对结构风险最小化（SRM）策略的考虑不足。为了解决这一限制，本文提出了一种物理信息的连续时间强化学习（PICRL），从ERM和SRM的角度验证模型的有效性。具体来说，我们从理论上分析SRM在强化学习模型中的机制开始。然后，将物理信息集成到离散和连续强化学习算法中进行对比实验。最后，我们系统地研究了各种物理信息和边界约束对这两种学习框架的影响。在PandaGym上的实验结果表明，该方法在离散时间和连续时间强化学习框架中都取得了相当或更好的性能。这为其在小时间间隔动态系统的控制策略学习方面的显著优势提供了强有力的证据。

{"title":"Physics-informed continuous-time reinforcement learning with data-driven approach for robotic arm manipulation","authors":"Jin-Qiang Wang , Lirong Song , Jun Shen , Binbin Yong , Xiaoteng Han , Yuanbo Jiang , Mona Raoufi , Qingguo Zhou","doi":"10.1016/j.jii.2025.101008","DOIUrl":"10.1016/j.jii.2025.101008","url":null,"abstract":"<div><div>Deep reinforcement learning (DRL) plays a crucial role in complex sequential decision-making tasks. However, existing data-driven DRL methods primarily rely on an empirical risk minimization (ERM) strategy to fit optimal value function models. This approach often neglects the environment’s dynamical system properties, which in turn leads to an inadequate consideration of the structural risk minimization (SRM) strategy. To address this limitation, this paper proposes a physics-informed continuous-time reinforcement learning (PICRL) to validate model effectiveness from both ERM and SRM perspectives. Specifically, we begin by theoretically analyzing the mechanism of SRM in reinforcement learning models. Then, physics information is integrated into both discrete and continuous reinforcement learning algorithms for comparative experiments. Finally, we systematically examine the effects of various physics-informed and boundary constraints on these two learning frameworks. Experimental results on the PandaGym demonstrate that the proposed method achieves comparable or superior performance in both discrete and continuous-time reinforcement learning frameworks. This provides strong evidence for its significant advantages in learning control policies for dynamical systems with small time intervals.</div></div>","PeriodicalId":55975,"journal":{"name":"Journal of Industrial Information Integration","volume":"49 ","pages":"Article 101008"},"PeriodicalIF":10.4,"publicationDate":"2025-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145492241","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

From brain to reflex: An emergency response control architecture for embodied intelligent robots 从大脑到反射：嵌入式智能机器人的应急响应控制体系结构

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

Journal of Industrial Information Integration

Pub Date : 2025-11-10 DOI: 10.1016/j.jii.2025.101010

Cheng Wang , Shiyong Wang , Wujie Zhang , Min Xia , Zhenfeng Shi

The perception-control-execution layered architecture, commonly used in current embodied intelligent robot control systems, suffers from inherent latency caused by its serial processing mechanism, which limits a robot's ability to respond to sudden disturbances, such as falls and collisions. To overcome this bottleneck, this study proposes a biomimetic emergency response control architecture for embodied intelligent robots. This architecture is inspired by the collaborative control principles of higher-level central control and spinal reflex mechanisms in the human nervous system. In addition, this architecture decouples the process of conventional decision-making and planning from emergency response mechanisms, thus constructing a four-layer heterogeneous control framework containing a perception-planning layer, a motion control layer, an emergency response layer, and a physical execution layer. The perception-planning layer is responsible for scene understanding and long-term planning. The motion control layer performs precise control of the entire body's posture and motion trajectory. The emergency response layer transmits upper-layer control commands under normal conditions, achieving fine motion control. In the event of sudden disturbances, the emergency response layer receives sensor signals directly, without waiting for the perception and decision results of the perception-planning layer. A lightweight, online-learnable reflex rule base, such as a balance compensation mechanism based on contact force mutation thresholds, enables rapid response to sudden disturbances. The emergency response layer is used as an independent module in the embodied intelligent control architecture, addressing the serial delay problem and offering an innovative solution for improving motion robustness and operational safety of robots in highly dynamic and uncertain environments.

当前嵌入式智能机器人控制系统中常用的感知-控制-执行分层结构，由于其串行处理机制导致的固有延迟，限制了机器人对突发干扰（如跌倒和碰撞）的响应能力。为了克服这一瓶颈，本研究提出了一种具身智能机器人的仿生应急响应控制体系结构。这种结构的灵感来自于人类神经系统中高级中枢控制和脊柱反射机制的协同控制原理。此外，该体系结构将传统的决策和规划过程与应急响应机制解耦，从而构建了一个包含感知规划层、运动控制层、应急响应层和物理执行层的四层异构控制框架。感知规划层负责场景理解和长期规划。运动控制层对整个身体的姿态和运动轨迹进行精确控制。应急响应层在正常情况下传输上层控制命令，实现精细运动控制。在突发干扰情况下，应急响应层直接接收传感器信号，无需等待感知规划层的感知和决策结果。一个轻量级的，在线可学习的反射规则库，如基于接触力突变阈值的平衡补偿机制，可以快速响应突然的干扰。应急响应层作为嵌入式智能控制体系结构中的独立模块，解决了串行延迟问题，为提高机器人在高动态和不确定环境中的运动鲁棒性和运行安全性提供了创新的解决方案。

{"title":"From brain to reflex: An emergency response control architecture for embodied intelligent robots","authors":"Cheng Wang , Shiyong Wang , Wujie Zhang , Min Xia , Zhenfeng Shi","doi":"10.1016/j.jii.2025.101010","DOIUrl":"10.1016/j.jii.2025.101010","url":null,"abstract":"<div><div>The perception-control-execution layered architecture, commonly used in current embodied intelligent robot control systems, suffers from inherent latency caused by its serial processing mechanism, which limits a robot's ability to respond to sudden disturbances, such as falls and collisions. To overcome this bottleneck, this study proposes a biomimetic emergency response control architecture for embodied intelligent robots. This architecture is inspired by the collaborative control principles of higher-level central control and spinal reflex mechanisms in the human nervous system. In addition, this architecture decouples the process of conventional decision-making and planning from emergency response mechanisms, thus constructing a four-layer heterogeneous control framework containing a perception-planning layer, a motion control layer, an emergency response layer, and a physical execution layer. The perception-planning layer is responsible for scene understanding and long-term planning. The motion control layer performs precise control of the entire body's posture and motion trajectory. The emergency response layer transmits upper-layer control commands under normal conditions, achieving fine motion control. In the event of sudden disturbances, the emergency response layer receives sensor signals directly, without waiting for the perception and decision results of the perception-planning layer. A lightweight, online-learnable reflex rule base, such as a balance compensation mechanism based on contact force mutation thresholds, enables rapid response to sudden disturbances. The emergency response layer is used as an independent module in the embodied intelligent control architecture, addressing the serial delay problem and offering an innovative solution for improving motion robustness and operational safety of robots in highly dynamic and uncertain environments.</div></div>","PeriodicalId":55975,"journal":{"name":"Journal of Industrial Information Integration","volume":"49 ","pages":"Article 101010"},"PeriodicalIF":10.4,"publicationDate":"2025-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145485560","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 embodied intelligence-based online optimization methodology for injection molding process using multi-cavity hot-runner 多型腔热流道注射成型工艺的具体智能在线优化方法

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

Journal of Industrial Information Integration

Pub Date : 2025-11-09 DOI: 10.1016/j.jii.2025.101009

Hongyi Qu , Luo Fang , Jinbiao Tan

In the process of multi-cavity hot runner injection molding, the issue of mold filling imbalance caused by uneven temperature distribution significantly affects the quality of precision products such as optical lenses. Traditional methods primarily rely on mold thermal structure design and lack dynamic optimization strategies aimed at product quality. This paper proposes an embodied intelligent online optimization method integrated with digital twin technology, which fundamentally overcomes the limitations of traditional fixed-temperature control and offline optimization by enabling dynamic, data-driven adjustment of process parameters. By utilizing real-time process information from sensor readings within a batch, along with product quality data obtained through machine vision inspection after each batch, and employing a ‘mutual feedback’ sharing mechanism for multi-cavity process information, a ‘time-batch’ dual-scale real-time iterative learning and updating framework is established for the digital twin model. This approach enables closed-loop adaptive optimization of the mold filling state. Experimental results show that this method significantly outperforms traditional fixed temperature setting controls in terms of profile accuracy, offering an innovative solution for high-precision injection molding.

在多型腔热流道注射成型过程中，由于温度分布不均匀导致的充模不平衡问题严重影响光学透镜等精密产品的质量。传统方法主要依靠模具热结构设计，缺乏针对产品质量的动态优化策略。本文提出了一种结合数字孪生技术的嵌入式智能在线优化方法，通过实现工艺参数的动态、数据驱动调整，从根本上克服了传统的定温控制和离线优化的局限性。通过利用一批内传感器读数的实时工艺信息，以及每批后通过机器视觉检测获得的产品质量数据，并采用多腔工艺信息的“互反馈”共享机制，为数字孪生模型建立了“时间批”双尺度实时迭代学习和更新框架。该方法实现了充型状态的闭环自适应优化。实验结果表明，该方法在轮廓精度方面明显优于传统的固定温度设定控制，为高精度注射成型提供了一种创新的解决方案。

{"title":"An embodied intelligence-based online optimization methodology for injection molding process using multi-cavity hot-runner","authors":"Hongyi Qu , Luo Fang , Jinbiao Tan","doi":"10.1016/j.jii.2025.101009","DOIUrl":"10.1016/j.jii.2025.101009","url":null,"abstract":"<div><div>In the process of multi-cavity hot runner injection molding, the issue of mold filling imbalance caused by uneven temperature distribution significantly affects the quality of precision products such as optical lenses. Traditional methods primarily rely on mold thermal structure design and lack dynamic optimization strategies aimed at product quality. <em>This paper proposes an embodied intelligent online optimization method integrated with digital twin technology, which fundamentally overcomes the limitations of traditional fixed-temperature control and offline optimization by enabling dynamic, data-driven adjustment of process parameters</em>. By utilizing real-time process information from sensor readings within a batch, along with product quality data obtained through machine vision inspection after each batch, and employing a ‘mutual feedback’ sharing mechanism for multi-cavity process information, a ‘time-batch’ dual-scale real-time iterative learning and updating framework is established for the digital twin model. This approach enables closed-loop adaptive optimization of the mold filling state. Experimental results show that this method significantly outperforms traditional fixed temperature setting controls in terms of profile accuracy, offering an innovative solution for high-precision injection molding.</div></div>","PeriodicalId":55975,"journal":{"name":"Journal of Industrial Information Integration","volume":"49 ","pages":"Article 101009"},"PeriodicalIF":10.4,"publicationDate":"2025-11-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145473276","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 graph-embedded musical chairs optimization with secure elliptic encryption framework for intelligent edge computing in healthcare iot networks 基于安全椭圆加密框架的新型嵌入式音乐椅优化，用于医疗物联网智能边缘计算

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

Journal of Industrial Information Integration

Pub Date : 2025-11-05 DOI: 10.1016/j.jii.2025.101007

R. Gowthamani , S. Oswalt Manoj

Internet of Things (IoT) health systems have severe issues distinguishing malicious from legitimate traffic and ensuring secure and efficient data transmission for real-time patient care. Existing solutions have high complexity, low dynamic attack adaptability, and low encryption strength. For the purpose of solving these problems, this study suggests a security-enhanced intelligent edge computing system that involves Normalized Distance-Based Encoding (NDBE) for effective feature extraction, Adaptive Layout Decomposition with Graph Embedding Neural Networks (ADGENN) for malicious data identification, Musical Chairs Optimization Algorithm (MCOA) for adaptive hyperparameter tuning, and a novel Light-weight Dynamic Elliptic Curve Cryptography with Schoof's Algorithm (LDECCSA) for data encryption protection. Together, these modules enhance classification efficiency, reduce computational costs, and facilitate low-latency, safe communication. Evaluated on the ToN-IoT and CICIoMT2024 dataset, the system achieves up to 99.87 % accuracy, 97 % throughput, and a low latency of 1.2 s, which performs better than current cutting-edge solutions by a large margin. The significance of this work is that it has the capacity to handle some of the most significant issues in healthcare. Systems are currently confronting, wherein IoT devices and edge computing have taken patient tracking to a new height, but also created gargantuan challenges such as cyberattacks, data breaches, and performance congestion. The major novelties are the application of NDBE for pre-processing network traffic, dynamic graph-based classification through ADGENN, resource-aware optimization through MCOA, and light-weighted, secure ECC with dynamic curve generation. While the model shows better efficiency and resilience, its dependence on pre-labeled datasets might restrict flexibility towards unknown real-world threats, and resource-limited IoT devices might struggle with heavy computation. In summary, the framework offers a real-world, scalable solution for real-time threat identification, secure data transfer, and effective healthcare surveillance in an IoT-based, cutting-edge healthcare environment.

物联网（IoT）卫生系统在区分恶意流量和合法流量以及确保安全高效的数据传输以实现实时患者护理方面存在严重问题。现有的解决方案存在复杂度高、动态攻击适应性差、加密强度低等问题。为了解决这些问题，本研究提出了一种安全增强的智能边缘计算系统，该系统包括用于有效特征提取的归一化距离编码（NDBE）、用于恶意数据识别的基于图嵌入神经网络的自适应布局分解（ADGENN）、用于自适应超参数调优的音乐椅子优化算法（MCOA）、用于自适应超参数调优的智能边缘计算系统。基于Schoof算法的轻型动态椭圆曲线加密（LDECCSA）数据加密保护。这些模块共同提高了分类效率，降低了计算成本，并促进了低延迟、安全的通信。在ToN-IoT和CICIoMT2024数据集上进行评估，该系统达到99.87%的准确率、97%的吞吐量和1.2 s的低延迟，大大优于当前的前沿解决方案。这项工作的意义在于，它有能力处理医疗保健中一些最重要的问题。系统目前面临的问题是，物联网设备和边缘计算将患者跟踪带到了一个新的高度，但也带来了巨大的挑战，如网络攻击、数据泄露和性能拥堵。主要的创新点是应用NDBE对网络流量进行预处理，通过ADGENN进行基于动态图的分类，通过MCOA进行资源感知优化，以及采用动态曲线生成的轻量级安全ECC。虽然该模型显示出更好的效率和弹性，但它对预标记数据集的依赖可能会限制对未知现实世界威胁的灵活性，并且资源有限的物联网设备可能会在繁重的计算中挣扎。总之，该框架为基于物联网的尖端医疗保健环境中的实时威胁识别、安全数据传输和有效医疗保健监控提供了一个真实的、可扩展的解决方案。

{"title":"A Novel graph-embedded musical chairs optimization with secure elliptic encryption framework for intelligent edge computing in healthcare iot networks","authors":"R. Gowthamani , S. Oswalt Manoj","doi":"10.1016/j.jii.2025.101007","DOIUrl":"10.1016/j.jii.2025.101007","url":null,"abstract":"<div><div>Internet of Things (IoT) health systems have severe issues distinguishing malicious from legitimate traffic and ensuring secure and efficient data transmission for real-time patient care. Existing solutions have high complexity, low dynamic attack adaptability, and low encryption strength. For the purpose of solving these problems, this study suggests a security-enhanced intelligent edge computing system that involves Normalized Distance-Based Encoding (NDBE) for effective feature extraction, Adaptive Layout Decomposition with Graph Embedding Neural Networks (ADGENN) for malicious data identification, Musical Chairs Optimization Algorithm (MCOA) for adaptive hyperparameter tuning, and a novel Light-weight Dynamic Elliptic Curve Cryptography with Schoof's Algorithm (LDECCSA) for data encryption protection. Together, these modules enhance classification efficiency, reduce computational costs, and facilitate low-latency, safe communication. Evaluated on the ToN-IoT and CICIoMT2024 dataset, the system achieves up to 99.87 % accuracy, 97 % throughput, and a low latency of 1.2 s, which performs better than current cutting-edge solutions by a large margin. The significance of this work is that it has the capacity to handle some of the most significant issues in healthcare. Systems are currently confronting, wherein IoT devices and edge computing have taken patient tracking to a new height, but also created gargantuan challenges such as cyberattacks, data breaches, and performance congestion. The major novelties are the application of NDBE for pre-processing network traffic, dynamic graph-based classification through ADGENN, resource-aware optimization through MCOA, and light-weighted, secure ECC with dynamic curve generation. While the model shows better efficiency and resilience, its dependence on pre-labeled datasets might restrict flexibility towards unknown real-world threats, and resource-limited IoT devices might struggle with heavy computation. In summary, the framework offers a real-world, scalable solution for real-time threat identification, secure data transfer, and effective healthcare surveillance in an IoT-based, cutting-edge healthcare environment.</div></div>","PeriodicalId":55975,"journal":{"name":"Journal of Industrial Information Integration","volume":"49 ","pages":"Article 101007"},"PeriodicalIF":10.4,"publicationDate":"2025-11-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145447594","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

LightPose: A lightweight fatigue-aware pose estimation framework LightPose：一个轻量级的疲劳感知姿态估计框架

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.2025.100988

Da Long, Sheng Yang

Fatigue assessment based on human motion plays a critical role in human-centric intelligent manufacturing, intelligent monitoring, and ergonomics. This growing demand underscores the need for low-cost, high-precision pose estimation techniques with broad application adaptability. To meet these requirements, we propose LightPose, a lightweight human pose estimation framework guided by bone segment principles. LightPose is designed to balance spatial accuracy with computational efficiency, delivering pose quality comparable to recent sequence-based baselines while remaining lightweight enough for real-time, fatigue-aware analysis. The framework incorporates a dual-stream supervision mechanism that enforces local geometric consistency through mutual prediction between joint pairs on the same bone segment. Additionally, kinematic constraints and fatigue-relevant metric regulations are embedded within the training objective, promoting biomechanical plausibility and alignment with fatigue-related motion patterns. Experimental results on standard 3D pose estimation benchmarks demonstrate that LightPose delivers competitive accuracy with reduced computational cost. Further evaluations confirm its effectiveness in estimating fatigue-related kinematic indicators, establishing its suitability for fatigue detection tasks. By effectively bridging efficiency and biomechanical relevance, LightPose presents a promising front-end solution for fatigue-aware motion analysis in manufacturing settings.

基于人体运动的疲劳评估在以人为中心的智能制造、智能监控和人机工程学中具有重要作用。这种不断增长的需求强调了对具有广泛应用适应性的低成本、高精度姿态估计技术的需求。为了满足这些要求，我们提出了LightPose，这是一个基于骨段原理的轻量级人体姿态估计框架。LightPose旨在平衡空间精度和计算效率，提供与最近基于序列的基线相媲美的姿态质量，同时保持足够轻量的实时疲劳感知分析。该框架采用双流监督机制，通过同一骨段上关节对之间的相互预测来强制局部几何一致性。此外，运动学约束和疲劳相关的度量规则嵌入到训练目标中，促进生物力学的合理性，并与疲劳相关的运动模式对齐。在标准3D姿态估计基准上的实验结果表明，LightPose在降低计算成本的同时提供了具有竞争力的精度。进一步的评估证实了它在估计疲劳相关运动学指标方面的有效性，并确定了它对疲劳检测任务的适用性。通过有效地连接效率和生物力学相关性，LightPose为制造环境中的疲劳感知运动分析提供了一个有前途的前端解决方案。

{"title":"LightPose: A lightweight fatigue-aware pose estimation framework","authors":"Da Long, Sheng Yang","doi":"10.1016/j.jii.2025.100988","DOIUrl":"10.1016/j.jii.2025.100988","url":null,"abstract":"<div><div>Fatigue assessment based on human motion plays a critical role in human-centric intelligent manufacturing, intelligent monitoring, and ergonomics. This growing demand underscores the need for low-cost, high-precision pose estimation techniques with broad application adaptability. To meet these requirements, we propose <em>LightPose</em>, a lightweight human pose estimation framework guided by bone segment principles. <em>LightPose</em> is designed to balance spatial accuracy with computational efficiency, delivering pose quality comparable to recent sequence-based baselines while remaining lightweight enough for real-time, fatigue-aware analysis. The framework incorporates a dual-stream supervision mechanism that enforces local geometric consistency through mutual prediction between joint pairs on the same bone segment. Additionally, kinematic constraints and fatigue-relevant metric regulations are embedded within the training objective, promoting biomechanical plausibility and alignment with fatigue-related motion patterns. Experimental results on standard 3D pose estimation benchmarks demonstrate that <em>LightPose</em> delivers competitive accuracy with reduced computational cost. Further evaluations confirm its effectiveness in estimating fatigue-related kinematic indicators, establishing its suitability for fatigue detection tasks. By effectively bridging efficiency and biomechanical relevance, <em>LightPose</em> presents a promising front-end solution for fatigue-aware motion analysis in manufacturing settings.</div></div>","PeriodicalId":55975,"journal":{"name":"Journal of Industrial Information Integration","volume":"48 ","pages":"Article 100988"},"PeriodicalIF":10.4,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145383881","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

Bi-objective sustainable urban logistics vehicle routing problem with workload balance 具有负载平衡的双目标可持续城市物流车辆路径问题

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.2025.100985

Wenyan Zhao, Yaguang Yuan, Cong Cheng, Wenheng Liu

The rapid advancement of e-commerce has driven unprecedented expansion in urban logistics networks, where their sustainability is constrained by multifaceted factors including strict time-bound service requirements, employee’s satisfaction, traffic congestion, and carbon emission regulations. Among these critical elements, employee’s satisfaction reflected by the workload balance not only influences task execution quality but also affects long-term operational sustainability for logistics enterprises, rendering its enhancement an urgent priority in contemporary urban logistics practices. This paper thus investigates a sustainable urban logistics vehicle routing problem mainly focusing on this perspective. Initially, a bi-objective mixed-integer programming model is formulated to simultaneously minimize total delivery cost and workload balance. Subsequently, a hybrid metaheuristic algorithm combining path relinking (PR) with multi-directional local search framework is developed. The adaptive large neighborhood search is adopted to facilitate the intensive local exploration, while PR techniques enhance global search capabilities through systematic solution space diversification. The algorithm's validity is rigorously verified through comparative analyses with state of art multi-objective optimization algorithms using adapted benchmark instances. Computational results demonstrate the algorithmic effectiveness and efficiency, accompanied by detailed analyses of approximate Pareto front and model’s sensitivity. These findings advance the field of urban delivery and provide practical insights for implementing efficient and sustainable urban logistic systems.

电子商务的快速发展推动了城市物流网络的空前扩张，其可持续性受到多方面因素的制约，包括严格的限时服务要求、员工满意度、交通拥堵和碳排放法规。在这些关键要素中，以工作量平衡为体现的员工满意度不仅影响任务的执行质量，而且影响物流企业的长期运营可持续性，因此提高员工满意度是当代城市物流实践的当务之急。因此，本文主要围绕这一视角研究可持续城市物流车辆路径问题。首先，建立了一个双目标混合整数规划模型，同时最小化总交付成本和工作量平衡。随后，提出了一种结合路径链接和多向局部搜索框架的混合元启发式算法。采用自适应大邻域搜索，增强局部密集搜索能力，PR技术通过系统解空间多样化增强全局搜索能力。通过与多目标优化算法的对比分析，验证了该算法的有效性。计算结果表明了算法的有效性和有效性，并详细分析了近似帕累托前沿和模型的灵敏度。这些发现推动了城市物流领域的发展，并为实施高效和可持续的城市物流系统提供了实际的见解。

{"title":"Bi-objective sustainable urban logistics vehicle routing problem with workload balance","authors":"Wenyan Zhao, Yaguang Yuan, Cong Cheng, Wenheng Liu","doi":"10.1016/j.jii.2025.100985","DOIUrl":"10.1016/j.jii.2025.100985","url":null,"abstract":"<div><div>The rapid advancement of e-commerce has driven unprecedented expansion in urban logistics networks, where their sustainability is constrained by multifaceted factors including strict time-bound service requirements, employee’s satisfaction, traffic congestion, and carbon emission regulations. Among these critical elements, employee’s satisfaction reflected by the workload balance not only influences task execution quality but also affects long-term operational sustainability for logistics enterprises, rendering its enhancement an urgent priority in contemporary urban logistics practices. This paper thus investigates a sustainable urban logistics vehicle routing problem mainly focusing on this perspective. Initially, a bi-objective mixed-integer programming model is formulated to simultaneously minimize total delivery cost and workload balance. Subsequently, a hybrid metaheuristic algorithm combining path relinking (PR) with multi-directional local search framework is developed. The adaptive large neighborhood search is adopted to facilitate the intensive local exploration, while PR techniques enhance global search capabilities through systematic solution space diversification. The algorithm's validity is rigorously verified through comparative analyses with state of art multi-objective optimization algorithms using adapted benchmark instances. Computational results demonstrate the algorithmic effectiveness and efficiency, accompanied by detailed analyses of approximate Pareto front and model’s sensitivity. These findings advance the field of urban delivery and provide practical insights for implementing efficient and sustainable urban logistic systems.</div></div>","PeriodicalId":55975,"journal":{"name":"Journal of Industrial Information Integration","volume":"48 ","pages":"Article 100985"},"PeriodicalIF":10.4,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145416550","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

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