A Digital Twin and Knowledge Graph Fusion Framework for Industrial Intelligence

A Digital Twin and Knowledge Graph Fusion Framework for Industrial Intelligence | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article A Digital Twin and Knowledge Graph Fusion Framework for Industrial Intelligence Jia Li, Jingwei Zhao, Xuwei Huang, Ruiqi Li This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8401810/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract In modern industrial systems, Digital Twins (DT) focus on mapping physical states, simulating conditions, and predicting operations. Knowledge Graphs (KG) specialize in structured knowledge representation, cross-entity association, and logical inference. Despite their complementarity, the separation between the data and semantic layers leads to challenges such as limited model reuse, weak semantic interpretation, and a lack of closed-loop control. This paper presents a four-layer DT–KG fusion architecture (Q-Layer) that defines collaborative mechanisms from entity mapping through cognitive decision-making and distills four fusion patterns suited to diverse industrial scenarios. Q-Layer addresses the problem of heterogeneous data silos by establishing a closed loop of perception, semantics, inference, and decision, significantly enhancing system intelligence and adaptability. This framework offers a scalable pathway and theoretical foundation for advancing industrial intelligence. Digital Twins Knowledge Graphs Q-Layer four fusion patterns Figures Figure 1 Figure 2 Figure 3 1. Introduction Against the backdrop of the rapid advancement of global intelligent manufacturing systems, DT technology has enabled transparent visualization of production processes and predictive analysis of operational behaviours by establishing a high-fidelity, bidirectional, real-time interaction mechanism between physical entities and virtual models. However, the current construction of DT systems remains primarily focused on modelling, simulation, and data-driven prediction, exhibiting inherent limitations in representing cross-system knowledge and modelling complex semantic associations [ ¹ ] . As a result, DT systems face significant constraints in cognitive-level application scenarios such as mechanism interpretation, root cause diagnosis of anomalies, and optimization decision-making, due to the lack of structured knowledge support. KG, with its robust capabilities in ontology modelling and relational organization, offers a machine-computable knowledge network that provides a solid semantic foundation for the explicit representation of industrial mechanisms, rule constraints, and experiential associations [ 2 – 3 ] . This paper argues that the deep integration of DT and KG is not merely a complementary interaction at the information layer, but rather a pivotal technological enabler for the evolution of industrial intelligence from "operational perception" to "cognitive decision-making." To this end, the paper proposes the Q-Layer fusion architecture, aiming to establish a reusable and scalable knowledge-enhanced twin system framework to address the prevalent issues of "knowledge scarcity" and "closed-loop disconnection" in current industrial DT systems. 2. Theoretical Foundations and Synergistic Evolution Logic of the Fusion Technology 2.1 Analysis of Data Characteristics and Semantic Limitations of DT DT systems focus on the dynamic mapping of physical entities, enabling continuous and real-time generation of massive time-series data that describe operational states, environmental boundaries, and behavioural sequences. However, in their native form, these data primarily exist as raw numerical values, lacking inherent semantic context and logical structure [ 4 ] . As a result, they are insufficient to directly support advanced inference tasks such as Failure Mode and Effects Analysis (FMEA) and process parameter optimization. The underlying issue lies in the fact that, although DT data exhibit high fidelity and real-time characteristics, they lack knowledge-oriented representation and logical relational structures, which hinders the direct application of knowledge engineering [ 5 ] . 2.2 Logical Empowerment of Industrial Systems through KG KG, with ontology as the core abstraction of knowledge, explicitly represent the attributes, structural relationships, and operational mechanisms of industrial entities through a triplet structure (Entity, Relation, Entity). This graph-based structure can simultaneously accommodate static manufacturing standards, equipment specifications, and dynamic operational knowledge [ 6 ] . The logical expressiveness of the KG enables the injection of semantic context into the real-time data of DT systems, endowing the system with capabilities for causal inference and cognitive interpretation. This effectively addresses the inherent limitations of purely data-driven or physics-based simulation models in terms of interpretability [ 1 , 3 ] . 2.3 A Dual-Driven Integration Paradigm of “Perception and Cognition.” The collaborative mechanism between DT and KG can be precisely characterized as: data-driven perception and knowledge-driven cognition. As the foundational data source, the DT system continuously streams real-time operational data into the KG, enabling the knowledge network to evolve dynamically in tandem with the physical system's state [ 7 ] . Conversely, the KG, leveraging its built-in inference mechanisms, drives the DT model in reverse to perform parameter adjustments and behavioural optimization, thereby enabling a closed-loop flow from perceptual information to cognitive feedback. This bidirectional coupling serves as the core driving force for the adaptive evolution of industrial intelligence systems [ 8 ] . 3. Q-Layer: A Four-Layer Architecture for Deep Integration of DT and KG To establish a unified semantic space and a coordinated operational mechanism between DT and KG, this paper proposes the Q-Layer integration architecture(see Fig. 1 .). Through a hierarchical design, it enables end-to-end integration of a unified identification system, a semantic relationship network, a dynamic state injection mechanism, and an intelligent inference loop. 3.1 Entity Mapping Layer: Semantic-Physical Object As the foundational layer of the architecture, this component achieves consistent semantic representation between heterogeneous industrial objects and knowledge-based semantic entities. It uses Uniform Resource Identifiers (URI) or Globally Unique Identifiers (GUID) to map device components, sensor variables, and virtual simulation parameters to ontological entities. Previously fragmented data from production stages—such as human, machine, material, method, and environment—are now efficiently encapsulated as semantically interpretable objects. This process eliminates structural data silos and establishes a foundation for cross-system interoperability [ 2 , 5 , 9 ] . 3.2 Semantic Association Layer: Construction of Topological Networks and Process Logic Chains The Semantic Association Layer is designed to establish complex relational logic among entities and serves as a critical component in structuring the knowledge network. This layer explicitly organizes the structural topology of equipment and the causal logic of processes by introducing a rich set of relational predicates (e.g.hasComponent, partOf, interactWith). Isolated entities are structured using the Resource Description Framework (RDF), enabling the construction of a global knowledge topology. For instance, when a local anomaly is detected—such as increased bearing vibration—the system can perform rapid multi-hop tracing along semantic links to accurately identify the potential impact scope and upstream root causes of the fault. This facilitates a significant advancement from isolated state perception to comprehensive semantic insight [ 4 , 10 ] . 3.3 State Synchronization Layer: Real-Time Semantic Injection of Data and Event Evolution This layer is responsible for processing high-frequency dynamic data streams from the DT system and ensuring high real-time synchronization between the knowledge network and the physical state. By leveraging high-performance data buses such as Kafka or MQTT to establish an event-driven mechanism, time-series data collected in real time is written into the KG. During this process, the system transforms "raw numerical values" into "semantic events": indicators such as temperature and vibration are no longer meaningless floating-point numbers, but are instead treated as real-time attribute nodes that trigger updates to the knowledge network. This ensures the synchronized evolution and timeliness of the knowledge network in alignment with the physical operational state [ 5 , 9 , 11 , 12 ] . 3.4 Cognitive Decision Layer: Realization of Inference-Driven Feedback Control Loop Positioned at the top of the architecture, the cognitive decision layer serves as the critical component for enabling system intelligence. This layer integrates an OWL-DL (Web Ontology Language Description Logics) reasoner and a SWRL (Semantic Web Rule Language) rule base to support in-depth analysis of operational conditions, derivation of optimization strategies, and implementation of closed-loop control. When the inference engine detects operational anomalies or identifies potential areas for optimization, the system can autonomously generate control strategies based on prior knowledge embedded in the KG—such as shutdown alerts, process parameter adjustments, or load switching. These strategies are then written back to the physical execution layer via the DT interface, thereby establishing an intelligent decision-making loop encompassing perception → inference → feedback → verification [ 5 , 14 ] . 4. In-Depth Refinement of Industrial Collaborative Integration Patterns Building upon the logical foundation of the Q-Layer architecture, this paper distills four representative DT–KG integrated operational patterns(see Fig. 2 .) to address the varying demands of industrial intelligence across different levels. 4.1 Semantics-Driven Asset Standardization Pattern This pattern adopts the domain ontology within the KG as the absolute core specification, with DT models regarded as concrete instantiations of this ontology in the physical world.The core value of this model lies in addressing the interoperability challenges among industrial equipment from multiple vendors, batches, and highly fragmented sources through a unified semantic ontology. By enforcing standardized semantics, it enables plug-and-play functionality for industrial assets, significantly reducing the integration cost of DT systems [ 15 , 16 , 17 ] . 4.2 Data-Evolving Dynamic Knowledge Model In this model, the primary role of the DT is to serve as a high-throughput data source, continuously feeding real-time and historical samples into KG. The research focus is on leveraging graph mining techniques and machine learning methods to automatically discover new correlation patterns, latent fault features, or previously unknown causal relationships from massive volumes of simulated data, thereby enabling the autonomous expansion and dynamic evolution of the knowledge base. This model offers significant advantages in scenarios that require the accumulation of operational experience, such as prognostics and health management (PHM) for long-lifecycle equipment [ 9 , 12 , 18 ] . 4.3 Rule-Constrained Mechanism-Enhanced Model In this model, structured representations of physical laws, industry standards, and safety operation protocols within the KG are treated as a priori knowledge to impose hard constraints on the simulation behaviour of the DT model. This approach effectively addresses the issue of non-physical deviations or "hallucinations" that may arise in purely data-driven models during prediction, ensuring that the DT’s outputs conform to physical principles and safety regulations. In scenarios with strong mechanistic dependencies, such as complex welding processes or chemical reactions, the injection of knowledge significantly enhances the reliability and robustness of the twin system [ 19 , 20 , 21 ] . 4.4 Cognition-Driven Closed-Loop Decision-Making Model This model represents the most advanced form of integration between DT and KG, emphasizing iterative feedback and cyclical interaction between the two. In this framework, perceptual data from the DT flows into the KG for logical inference. The resulting optimized decisions and control strategies are then fed back into the DT for validation and behavioural adjustment, forming a continuously evolving spiral loop. This cognitive closed-loop mechanism enables the system to dynamically adapt to environmental changes during operation, serving as a critical technological pathway toward achieving “unmanned production” and “intelligent flexible manufacturing.” [ 1 , 2 , 3 , 7 ] 5. Multi-Level Fusion Validation in Wind Power Operation and Maintenance Scenarios 5.1 Multi-Level Fusion Validation in Wind Power Operation and Maintenance Scenarios Taking the predictive maintenance of large-scale wind turbines as an example, the Q-Layer framework enables the construction of a unified entity mapping by semantically encapsulating the turbine structural model. The semantic association layer explicitly establishes complex causal chains among blade stress, generator load, and bearing failures. The state synchronization layer continuously injects high-frequency vibration and environmental data in real time. Finally, the cognitive decision-making layer derives the optimal pitch control strategy based on mechanistic rules and KG reasoning, and dispatches it for execution, thereby achieving real-time optimized control(see Fig. 3 .). Results indicate that this integrated process effectively reduces the rate of unplanned downtime failures and enhances overall power generation efficiency [ 22 , 23 , 24 ] . 5.2 Implementation Pathway and Engineering Feasibility Discussion The engineering implementation of the Q-Layer architecture is recommended to proceed in three stages: (1) Knowledge Engineering Stage. Develop a domain-specific industrial ontology system and entity mapping specifications, completing the structuring of domain knowledge and the construction of the KG. (2) Data Stream Integration Stage. Build a real-time streaming data pipeline that supports high throughput and low latency to ensure the semantic injection of twin data into the KG. (3) Inference Deployment Phase. A distributed inference system capable of operating in real-time industrial environments should be developed, leveraging edge computing capabilities to enable millisecond-level cognitive closed-loop responses. With the maturation of distributed graph database technologies, edge computing, and high-performance inference engines, the implementation of knowledge-enhanced DT systems with real-time feedback capabilities has become both technically feasible and practically achievable [ 25 ] . 6. Conclusions and Outlook This paper systematically proposes Q-Layer, a DT–KG integration framework tailored for industrial intelligence, along with four core collaborative patterns. Through rigorous academic analysis, the research demonstrates that knowledge injection is the fundamental approach to addressing the issues of semantic deficiency and opaque decision-making in DT systems. This paradigm of deep integration not only enables high-fidelity simulation of physical processes but also provides interpretable and traceable logical support for complex production decision-making [ 5 , 9 ] . The Q-Layer architecture offers a clear theoretical foundation and technical framework for building the next generation of adaptive and self-evolving industrial intelligence systems. Future research will focus on addressing key technical challenges, including large-scale real-time inference optimization, automated knowledge extraction mechanisms, and multimodal data-knowledge fusion, with the goal of developing industrial intelligence systems capable of autonomous learning and achieving a significant leap in the level of manufacturing intelligence [ 26 , 27 ] . Declarations Conflicts of interest/Competing interests All authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest or non-financial interest in the subject matter or materials discussed in this manuscript. Ethics approval Not applicable. (This manuscript does not contain studies with human participants or animals performed by any of the authors.) Consent to participate Not applicable. Consent for publication Not applicable. Funding No funds, grants, or other support was received for conducting this study. Author Contribution J.L. conceived the study and was responsible for drafting the manuscript. J.Z. and X.H. conducted the investigation and collation of the wind turbine operation and maintenance case studies, and produced Figures 1–3. R.L. participated in reviewing and editing the manuscript. All authors have read and approved the final version. Acknowledgements Not applicable. Data Availability Not applicable. (This manuscript does not report data generation or analysis. All data and materials supporting the findings are included within the article.) References Su C, Tang X, Jiang Q, Han Y, Wang T, Jiang D. Digital twin system for manufacturing processes based on a multi-layer knowledge graph model. Sci Rep. 2025;15. https://doi.org/10.1038/s41598-024-85053-0 . Su C, Tang X, Han Y, Wang T, Jiang D. Cognitive digital twin in manufacturing process: integrating the knowledge graph for enhanced human-centric Industry 5.0. Int J Prod Res. 2024;1–22. https://doi.org/10.1080/00207543.2024.2435583 . Su C, Han Y, Tang X, Jiang Q, Wang T, He Q. Knowledge-based digital twin system: Using a knowlege-driven approach for manufacturing process modeling. Computers Ind Volumes. 2024. 104101.https://doi.org/10.1016/j.compind.2024.104101 . 159–160,2024,. Karabulut E, Pileggi S, Groth P, Degeler V, ArXiv. abs/2308.15168. https://doi.org/10.1016/j.future.2023.12.013 Akroyd J, Mosbach S, Bhave A, Kraft M. Universal Digital Twin - A Dynamic Knowledge Graph. Data-Centric Eng. 2021. 10.1017/dce.2021.10 . 2. Lim K, Yosal T, Chen C, Zheng P, Wang L, Xu X. Graph-enabled cognitive digital twins for causal inference in maintenance processes. Int J Prod Res. 2023;62:4717–34. https://doi.org/10.1080/00207543.2023.2274335 . Feng J, Tang H, Zhou S, Cai Y, Zhang J. Cognitive Digital Twins of the natural environment: Framework and application. Eng Appl Artif Intell. 2025;139:109587. https://doi.org/10.1016/j.engappai.2024.109587 . Stavropoulou G, Tsitseklis K, Mavraidi L, Chang K, Zafeiropoulos A, Karyotis V, Papavassiliou S. Digital Twin Meets Knowledge Graph for Intelligent Manufacturing Processes. Sensors. 2024;24. https://doi.org/10.3390/s24082618 . Ramonell C, Chacón R, Posada H. Knowledge graph-based data integration system for digital twins of built assets. Autom Constr. 2023. https://doi.org/10.1016/j.autcon.2023.105109 . Jia M, Hu J, Liu Y, Gao Z, Yao Y, Industrial. & Engineering Chemistry Research. https://doi.org/10.1021/acs.iecr.2c03628 Hofmeister M, Bai J, Brownbridge G, Mosbach S, Lee K, Farazi F, Hillman M, Agarwal M, Ganguly S, Akroyd J, Kraft M. Semantic agent framework for automated flood assessment using dynamic knowledge graphs. Data-Centric Eng. 2024;5. 10.1017/dce.2024.11 . Mortlock T, Muthirayan D, Yu S, Khargonekar P, Faruque M. Graph Learning for Cognitive Digital Twins in Manufacturing Systems. IEEE Trans Emerg Top Comput. 2021;10:34–45. 10.1109/TETC.2021.3132251 . Villalonga A, Negri E, Biscardo G, Castaño F, Haber R, Fumagalli L, Macchi M. A decision-making framework for dynamic scheduling of cyber-physical production systems based on digital twins. Annu Rev Control. 2021;51:357–73. https://doi.org/10.1016/j.arcontrol.2021.04.008 . Chang L, Zhang L, Fu C, Chen Y. Transparent Digital Twin for Output Control Using Belief Rule Base. IEEE Trans Cybernetics. 2021;52:10364–78. 10.1109/TCYB.2021.3063285 . Huang Y, Dhouib S, Medinacelli L, Malenfant J. (2023). Semantic Interoperability of Digital Twins: Ontology-based Capability Checking in AAS Modeling Framework. 2023 IEEE 6th International Conference on Industrial Cyber-Physical Systems (ICPS), 1–8. 10.1109/ICPS58381.2023.10128003 Göppert A, Grahn L, Rachner J, Grunert D, Hort S, Schmitt R. Pipeline for ontology-based modeling and automated deployment of digital twins for planning and control of manufacturing systems. J Intell Manuf. 2021;34:2133–52. https://doi.org/10.1007/s10845-021-01860-6 . Dittler D, Frank P, Hildebrandt G, Peterson L, Jazdi N, Weyrich M, ArXiv. abs/2507.03553.https://doi.org/10.48550/arXiv.2507.03553 Booyse W, Wilke D, Heyns S. Deep digital twins for detection, diagnostics and prognostics. Mech Syst Signal Process. 2020;140:106612. https://doi.org/10.1016/j.ymssp.2019.106612 . Bevilacqua M, Bottani E, Ciarapica F, Costantino F, Di Donato L, Ferraro A, Mazzuto G, Monteriù A, Nardini G, Ortenzi M, Paroncini M, Pirozzi M, Prist M, Quatrini E, Tronci M, Vignali G. Digital Twin Reference Model Development to Prevent Operators’ Risk in Process Plants. Sustainability. 2020;12:1088. https://doi.org/10.3390/su12031088 . Somers R, Douthwaite J, Wagg D, Walkinshaw N, Hierons R. Digital-twin-based testing for cyber-physical systems: A systematic literature review. Inf Softw Technol. 2022;156:107145. Omrany H, Al-Obaidi K, Husain A, Ghaffarianhoseini A. (2023). Digital Twins in the Construction Industry: A Comprehensive Review of Current Implementations, Enabling Technologies, and Future Directions. Sustainability. https://doi.org/10.1016/j.infsof.2022.107145 Xu T, Zhang X, Sun W, Wang B. Sensors. 2025;25. https://doi.org/10.3390/s25071972 . Intelligent Operation and Maintenance of Wind Turbines Gearboxes via Digital Twin and Multi-Source Data Fusion. Abdullahi I, Longo S, Samie M. (2024). Towards a Distributed Digital Twin Framework for Predictive Maintenance in Industrial Internet of Things (IIoT). Sensors (Basel, Switzerland), 24. https://doi.org/10.3390/s24082663 Van Dinter R, Tekinerdogan B, Catal C. Reference architecture for digital twin-based predictive maintenance systems. Comput Ind Eng. 2023;177:109099. https://doi.org/10.1016/j.cie.2023.109099 . Liu X, Wu X, He J, Gupta R. (2025). Intelligent Operation and Maintenance and Prediction Model Optimization for Improving Wind Power Generation Efficiency. 2025 7th International Congress on Human-Computer Interaction, Optimization and Robotic Applications (ICHORA), 1–7. https://doi.org/10.48550/arXiv.2506.16095 Ladj A, Wang Z, Meski O, Belkadi F, Ritou M, Da Cunha C. A knowledge-based Digital Shadow for machining industry in a Digital Twin perspective. J Manuf Syst. 2020. https://doi.org/10.1016/j.jmsy.2020.07.018 . Zheng P, Xia L, Li C, Li X, Liu B. Towards Self-X cognitive manufacturing network: An industrial knowledge graph-based multi-agent reinforcement learning approach. J Manuf Syst. 2021;61:16–26. https://doi.org/10.1016/j.jmsy.2021.08.002 . Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8401810","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":567010856,"identity":"feb39933-b675-4e14-bfa5-eafa7de3ed36","order_by":0,"name":"Jia Li","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAwUlEQVRIiWNgGAWjYBACPmaGhAMgBht7Y+ODD8RoYYNr4TncbDiDKC1wlkR6mzQHUVrYGR4e+FFjY88n+bBBmoHBTk63gQiHHew5lpbYJp3YYFzAkGxsdoAILQd4Gw4nsAG1JM9gOJC4jRgtB/82/LdnkzzYcJiHWC2HeRsOMLZJMDY2E69F5lhyYhtPYjPjDAMi/MLPfyb545saO3v59uPPf3yosJMjqIWBgScBiWNAUDkIsBM2dRSMglEwCkY4AACUmz7HmdGpkAAAAABJRU5ErkJggg==","orcid":"","institution":"Beijing University of Technology","correspondingAuthor":true,"prefix":"","firstName":"Jia","middleName":"","lastName":"Li","suffix":""},{"id":567010857,"identity":"84da09bb-ed15-4736-9307-38a4f3595297","order_by":1,"name":"Jingwei Zhao","email":"","orcid":"","institution":"Chongqing university of posts and telecommunications","correspondingAuthor":false,"prefix":"","firstName":"Jingwei","middleName":"","lastName":"Zhao","suffix":""},{"id":567010858,"identity":"a2ee5787-7995-4b75-ac35-d4caeb2b9627","order_by":2,"name":"Xuwei Huang","email":"","orcid":"","institution":"Chongqing university of posts and telecommunications","correspondingAuthor":false,"prefix":"","firstName":"Xuwei","middleName":"","lastName":"Huang","suffix":""},{"id":567010859,"identity":"ef7a8439-0d0a-4a63-b5f0-34a31d8af9b6","order_by":3,"name":"Ruiqi Li","email":"","orcid":"","institution":"China Electronics Standardization Institute","correspondingAuthor":false,"prefix":"","firstName":"Ruiqi","middleName":"","lastName":"Li","suffix":""}],"badges":[],"createdAt":"2025-12-19 07:23:43","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8401810/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8401810/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":99233353,"identity":"7d54552e-bc22-46f6-ba6b-ae432f666faf","added_by":"auto","created_at":"2025-12-30 12:41:08","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1027774,"visible":true,"origin":"","legend":"","description":"","filename":"ADigitalTwinandKnowledgeGraphFusionFrameworkforIndustrialIntelligence1223v1.docx","url":"https://assets-eu.researchsquare.com/files/rs-8401810/v1/666b38b2020de2e5bd00dcd9.docx"},{"id":99233358,"identity":"5f350eba-38c3-4555-b6e3-39d0f2fdeb2d","added_by":"auto","created_at":"2025-12-30 12:41:09","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5360,"visible":true,"origin":"","legend":"","description":"","filename":"9910225cfc3d4f668c8e958de9cf67c2.json","url":"https://assets-eu.researchsquare.com/files/rs-8401810/v1/ce40156327adf70a4569d06a.json"},{"id":99233361,"identity":"758f57da-51bf-4cc7-8625-fbd0420f7581","added_by":"auto","created_at":"2025-12-30 12:41:09","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":68783,"visible":true,"origin":"","legend":"","description":"","filename":"9910225cfc3d4f668c8e958de9cf67c21enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8401810/v1/bb61a40a076a8d3532da7343.xml"},{"id":99233394,"identity":"d8d8ade5-4ec4-4068-8845-202d2eaf3202","added_by":"auto","created_at":"2025-12-30 12:41:15","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":62339,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8401810/v1/ca886565214616df4e0b6350.png"},{"id":99233381,"identity":"23bdbb29-07cf-4f3a-854d-b70c2cdc0800","added_by":"auto","created_at":"2025-12-30 12:41:13","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":51840,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8401810/v1/294fbed6b48631490d3a9e15.png"},{"id":99233370,"identity":"69b992d9-41db-4906-924d-b1e6f72f5db2","added_by":"auto","created_at":"2025-12-30 12:41:11","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":50091,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8401810/v1/3b62e10c79f989d26f9278e9.png"},{"id":99233362,"identity":"674a15eb-c2bc-4473-9948-c80db77a7762","added_by":"auto","created_at":"2025-12-30 12:41:09","extension":"xml","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":68751,"visible":true,"origin":"","legend":"","description":"","filename":"9910225cfc3d4f668c8e958de9cf67c21structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8401810/v1/cc67b25dcdf44db87ae667d6.xml"},{"id":99233363,"identity":"3ad33bdd-06b9-4e9c-9c05-3e2e7829aab7","added_by":"auto","created_at":"2025-12-30 12:41:10","extension":"html","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":75436,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8401810/v1/5628a69569d93ec4fbbd1003.html"},{"id":99233351,"identity":"369e5f82-9215-40d6-9e74-a7890313a638","added_by":"auto","created_at":"2025-12-30 12:41:06","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":378311,"visible":true,"origin":"","legend":"\u003cp\u003eThe Q-Layer integration architecture\u003c/p\u003e","description":"","filename":"floatimage144.png","url":"https://assets-eu.researchsquare.com/files/rs-8401810/v1/2daceaf7b583b048811d4910.png"},{"id":99318064,"identity":"5afcfcbb-3d44-443a-b151-1b425e0de1bc","added_by":"auto","created_at":"2025-12-31 16:31:19","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":302711,"visible":true,"origin":"","legend":"\u003cp\u003eFour representative DT–KG integrated operational patterns\u003c/p\u003e","description":"","filename":"floatimage225.png","url":"https://assets-eu.researchsquare.com/files/rs-8401810/v1/7cc74fa2597291eba9d98400.png"},{"id":99233356,"identity":"3095321a-714e-42b2-a40f-ad44a91806a0","added_by":"auto","created_at":"2025-12-30 12:41:09","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":416968,"visible":true,"origin":"","legend":"\u003cp\u003eApplication case of the Q-Layer framework in predictive maintenance for large wind turbines\u003c/p\u003e","description":"","filename":"floatimage326.png","url":"https://assets-eu.researchsquare.com/files/rs-8401810/v1/9bf98265bde4368fe7b1ed3b.png"},{"id":100370810,"identity":"b2e6ed42-2df6-42fb-b9a1-a213c2e05aef","added_by":"auto","created_at":"2026-01-16 08:08:23","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1775898,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8401810/v1/6cdc3de9-55a4-471e-a9be-dddba8357444.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"A Digital Twin and Knowledge Graph Fusion Framework for Industrial Intelligence","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eAgainst the backdrop of the rapid advancement of global intelligent manufacturing systems, DT technology has enabled transparent visualization of production processes and predictive analysis of operational behaviours by establishing a high-fidelity, bidirectional, real-time interaction mechanism between physical entities and virtual models. However, the current construction of DT systems remains primarily focused on modelling, simulation, and data-driven prediction, exhibiting inherent limitations in representing cross-system knowledge and modelling complex semantic associations\u003csup\u003e[\u003c/sup\u003e\u0026sup1;\u003csup\u003e]\u003c/sup\u003e. As a result, DT systems face significant constraints in cognitive-level application scenarios such as mechanism interpretation, root cause diagnosis of anomalies, and optimization decision-making, due to the lack of structured knowledge support. KG, with its robust capabilities in ontology modelling and relational organization, offers a machine-computable knowledge network that provides a solid semantic foundation for the explicit representation of industrial mechanisms, rule constraints, and experiential associations\u003csup\u003e[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThis paper argues that the deep integration of DT and KG is not merely a complementary interaction at the information layer, but rather a pivotal technological enabler for the evolution of industrial intelligence from \"operational perception\" to \"cognitive decision-making.\" To this end, the paper proposes the Q-Layer fusion architecture, aiming to establish a reusable and scalable knowledge-enhanced twin system framework to address the prevalent issues of \"knowledge scarcity\" and \"closed-loop disconnection\" in current industrial DT systems.\u003c/p\u003e"},{"header":"2. Theoretical Foundations and Synergistic Evolution Logic of the Fusion Technology","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Analysis of Data Characteristics and Semantic Limitations of DT\u003c/h2\u003e \u003cp\u003eDT systems focus on the dynamic mapping of physical entities, enabling continuous and real-time generation of massive time-series data that describe operational states, environmental boundaries, and behavioural sequences. However, in their native form, these data primarily exist as raw numerical values, lacking inherent semantic context and logical structure\u003csup\u003e[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]\u003c/sup\u003e. As a result, they are insufficient to directly support advanced inference tasks such as Failure Mode and Effects Analysis (FMEA) and process parameter optimization. The underlying issue lies in the fact that, although DT data exhibit high fidelity and real-time characteristics, they lack knowledge-oriented representation and logical relational structures, which hinders the direct application of knowledge engineering\u003csup\u003e[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Logical Empowerment of Industrial Systems through KG\u003c/h2\u003e \u003cp\u003eKG, with ontology as the core abstraction of knowledge, explicitly represent the attributes, structural relationships, and operational mechanisms of industrial entities through a triplet structure (Entity, Relation, Entity). This graph-based structure can simultaneously accommodate static manufacturing standards, equipment specifications, and dynamic operational knowledge\u003csup\u003e[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]\u003c/sup\u003e. The logical expressiveness of the KG enables the injection of semantic context into the real-time data of DT systems, endowing the system with capabilities for causal inference and cognitive interpretation. This effectively addresses the inherent limitations of purely data-driven or physics-based simulation models in terms of interpretability\u003csup\u003e[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 A Dual-Driven Integration Paradigm of \u0026ldquo;Perception and Cognition.\u0026rdquo;\u003c/h2\u003e \u003cp\u003eThe collaborative mechanism between DT and KG can be precisely characterized as: data-driven perception and knowledge-driven cognition. As the foundational data source, the DT system continuously streams real-time operational data into the KG, enabling the knowledge network to evolve dynamically in tandem with the physical system's state\u003csup\u003e[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]\u003c/sup\u003e. Conversely, the KG, leveraging its built-in inference mechanisms, drives the DT model in reverse to perform parameter adjustments and behavioural optimization, thereby enabling a closed-loop flow from perceptual information to cognitive feedback. This bidirectional coupling serves as the core driving force for the adaptive evolution of industrial intelligence systems\u003csup\u003e[\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Q-Layer: A Four-Layer Architecture for Deep Integration of DT and KG","content":"\u003cp\u003eTo establish a unified semantic space and a coordinated operational mechanism between DT and KG, this paper proposes the Q-Layer integration architecture(see Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.). Through a hierarchical design, it enables end-to-end integration of a unified identification system, a semantic relationship network, a dynamic state injection mechanism, and an intelligent inference loop.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e3.1 Entity Mapping Layer: Semantic-Physical Object\u003c/h2\u003e \u003cp\u003eAs the foundational layer of the architecture, this component achieves consistent semantic representation between heterogeneous industrial objects and knowledge-based semantic entities. It uses Uniform Resource Identifiers (URI) or Globally Unique Identifiers (GUID) to map device components, sensor variables, and virtual simulation parameters to ontological entities. Previously fragmented data from production stages\u0026mdash;such as human, machine, material, method, and environment\u0026mdash;are now efficiently encapsulated as semantically interpretable objects. This process eliminates structural data silos and establishes a foundation for cross-system interoperability\u003csup\u003e[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Semantic Association Layer: Construction of Topological Networks and Process Logic Chains\u003c/h2\u003e \u003cp\u003eThe Semantic Association Layer is designed to establish complex relational logic among entities and serves as a critical component in structuring the knowledge network. This layer explicitly organizes the structural topology of equipment and the causal logic of processes by introducing a rich set of relational predicates (e.g.hasComponent, partOf, interactWith). Isolated entities are structured using the Resource Description Framework (RDF), enabling the construction of a global knowledge topology. For instance, when a local anomaly is detected\u0026mdash;such as increased bearing vibration\u0026mdash;the system can perform rapid multi-hop tracing along semantic links to accurately identify the potential impact scope and upstream root causes of the fault. This facilitates a significant advancement from isolated state perception to comprehensive semantic insight\u003csup\u003e[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e3.3 State Synchronization Layer: Real-Time Semantic Injection of Data and Event Evolution\u003c/h2\u003e \u003cp\u003eThis layer is responsible for processing high-frequency dynamic data streams from the DT system and ensuring high real-time synchronization between the knowledge network and the physical state. By leveraging high-performance data buses such as Kafka or MQTT to establish an event-driven mechanism, time-series data collected in real time is written into the KG. During this process, the system transforms \"raw numerical values\" into \"semantic events\": indicators such as temperature and vibration are no longer meaningless floating-point numbers, but are instead treated as real-time attribute nodes that trigger updates to the knowledge network. This ensures the synchronized evolution and timeliness of the knowledge network in alignment with the physical operational state\u003csup\u003e[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e3.4 Cognitive Decision Layer: Realization of Inference-Driven Feedback Control Loop\u003c/h2\u003e \u003cp\u003ePositioned at the top of the architecture, the cognitive decision layer serves as the critical component for enabling system intelligence. This layer integrates an OWL-DL (Web Ontology Language Description Logics) reasoner and a SWRL (Semantic Web Rule Language) rule base to support in-depth analysis of operational conditions, derivation of optimization strategies, and implementation of closed-loop control. When the inference engine detects operational anomalies or identifies potential areas for optimization, the system can autonomously generate control strategies based on prior knowledge embedded in the KG\u0026mdash;such as shutdown alerts, process parameter adjustments, or load switching. These strategies are then written back to the physical execution layer via the DT interface, thereby establishing an intelligent decision-making loop encompassing perception \u0026rarr; inference \u0026rarr; feedback \u0026rarr; verification\u003csup\u003e[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e"},{"header":"4. In-Depth Refinement of Industrial Collaborative Integration Patterns","content":"\u003cp\u003eBuilding upon the logical foundation of the Q-Layer architecture, this paper distills four representative DT\u0026ndash;KG integrated operational patterns(see Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e.) to address the varying demands of industrial intelligence across different levels.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e4.1 Semantics-Driven Asset Standardization Pattern\u003c/h2\u003e \u003cp\u003eThis pattern adopts the domain ontology within the KG as the absolute core specification, with DT models regarded as concrete instantiations of this ontology in the physical world.The core value of this model lies in addressing the interoperability challenges among industrial equipment from multiple vendors, batches, and highly fragmented sources through a unified semantic ontology. By enforcing standardized semantics, it enables plug-and-play functionality for industrial assets, significantly reducing the integration cost of DT systems\u003csup\u003e[\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e4.2 Data-Evolving Dynamic Knowledge Model\u003c/h2\u003e \u003cp\u003eIn this model, the primary role of the DT is to serve as a high-throughput data source, continuously feeding real-time and historical samples into KG. The research focus is on leveraging graph mining techniques and machine learning methods to automatically discover new correlation patterns, latent fault features, or previously unknown causal relationships from massive volumes of simulated data, thereby enabling the autonomous expansion and dynamic evolution of the knowledge base. This model offers significant advantages in scenarios that require the accumulation of operational experience, such as prognostics and health management (PHM) for long-lifecycle equipment\u003csup\u003e[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e4.3 Rule-Constrained Mechanism-Enhanced Model\u003c/h2\u003e \u003cp\u003eIn this model, structured representations of physical laws, industry standards, and safety operation protocols within the KG are treated as a priori knowledge to impose hard constraints on the simulation behaviour of the DT model. This approach effectively addresses the issue of non-physical deviations or \"hallucinations\" that may arise in purely data-driven models during prediction, ensuring that the DT\u0026rsquo;s outputs conform to physical principles and safety regulations. In scenarios with strong mechanistic dependencies, such as complex welding processes or chemical reactions, the injection of knowledge significantly enhances the reliability and robustness of the twin system\u003csup\u003e[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e4.4 Cognition-Driven Closed-Loop Decision-Making Model\u003c/h2\u003e \u003cp\u003eThis model represents the most advanced form of integration between DT and KG, emphasizing iterative feedback and cyclical interaction between the two. In this framework, perceptual data from the DT flows into the KG for logical inference. The resulting optimized decisions and control strategies are then fed back into the DT for validation and behavioural adjustment, forming a continuously evolving spiral loop. This cognitive closed-loop mechanism enables the system to dynamically adapt to environmental changes during operation, serving as a critical technological pathway toward achieving \u0026ldquo;unmanned production\u0026rdquo; and \u0026ldquo;intelligent flexible manufacturing.\u0026rdquo;\u003csup\u003e[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]\u003c/sup\u003e\u003c/p\u003e \u003c/div\u003e"},{"header":"5. Multi-Level Fusion Validation in Wind Power Operation and Maintenance Scenarios","content":"\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e5.1 Multi-Level Fusion Validation in Wind Power Operation and Maintenance Scenarios\u003c/h2\u003e \u003cp\u003eTaking the predictive maintenance of large-scale wind turbines as an example, the Q-Layer framework enables the construction of a unified entity mapping by semantically encapsulating the turbine structural model. The semantic association layer explicitly establishes complex causal chains among blade stress, generator load, and bearing failures. The state synchronization layer continuously injects high-frequency vibration and environmental data in real time. Finally, the cognitive decision-making layer derives the optimal pitch control strategy based on mechanistic rules and KG reasoning, and dispatches it for execution, thereby achieving real-time optimized control(see Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e.). Results indicate that this integrated process effectively reduces the rate of unplanned downtime failures and enhances overall power generation efficiency\u003csup\u003e[\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e5.2 Implementation Pathway and Engineering Feasibility Discussion\u003c/h2\u003e \u003cp\u003eThe engineering implementation of the Q-Layer architecture is recommended to proceed in three stages:\u003c/p\u003e \u003cp\u003e(1) Knowledge Engineering Stage. Develop a domain-specific industrial ontology system and entity mapping specifications, completing the structuring of domain knowledge and the construction of the KG.\u003c/p\u003e \u003cp\u003e(2) Data Stream Integration Stage. Build a real-time streaming data pipeline that supports high throughput and low latency to ensure the semantic injection of twin data into the KG.\u003c/p\u003e \u003cp\u003e(3) Inference Deployment Phase. A distributed inference system capable of operating in real-time industrial environments should be developed, leveraging edge computing capabilities to enable millisecond-level cognitive closed-loop responses.\u003c/p\u003e \u003cp\u003eWith the maturation of distributed graph database technologies, edge computing, and high-performance inference engines, the implementation of knowledge-enhanced DT systems with real-time feedback capabilities has become both technically feasible and practically achievable\u003csup\u003e[\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e"},{"header":"6. Conclusions and Outlook","content":"\u003cp\u003eThis paper systematically proposes Q-Layer, a DT\u0026ndash;KG integration framework tailored for industrial intelligence, along with four core collaborative patterns. Through rigorous academic analysis, the research demonstrates that knowledge injection is the fundamental approach to addressing the issues of semantic deficiency and opaque decision-making in DT systems. This paradigm of deep integration not only enables high-fidelity simulation of physical processes but also provides interpretable and traceable logical support for complex production decision-making\u003csup\u003e[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]\u003c/sup\u003e. The Q-Layer architecture offers a clear theoretical foundation and technical framework for building the next generation of adaptive and self-evolving industrial intelligence systems. Future research will focus on addressing key technical challenges, including large-scale real-time inference optimization, automated knowledge extraction mechanisms, and multimodal data-knowledge fusion, with the goal of developing industrial intelligence systems capable of autonomous learning and achieving a significant leap in the level of manufacturing intelligence\u003csup\u003e[\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eConflicts of interest/Competing interests\u003c/h2\u003e \u003cp\u003eAll authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest or non-financial interest in the subject matter or materials discussed in this manuscript.\u003c/p\u003e\u003cp\u003e \u003cstrong\u003eEthics approval\u003c/strong\u003e \u003cp\u003eNot applicable. (This manuscript does not contain studies with human participants or animals performed by any of the authors.)\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eConsent to participate\u003c/strong\u003e \u003cp\u003eNot applicable.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eConsent for publication\u003c/strong\u003e \u003cp\u003eNot applicable.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e \u003cp\u003eNo funds, grants, or other support was received for conducting this study.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eJ.L. conceived the study and was responsible for drafting the manuscript. J.Z. and X.H. conducted the investigation and collation of the wind turbine operation and maintenance case studies, and produced Figures 1\u0026ndash;3. R.L. participated in reviewing and editing the manuscript. All authors have read and approved the final version.\u003c/p\u003e\u003ch2\u003eAcknowledgements\u003c/h2\u003e \u003cp\u003eNot applicable.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eNot applicable. (This manuscript does not report data generation or analysis. All data and materials supporting the findings are included within the article.)\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eSu C, Tang X, Jiang Q, Han Y, Wang T, Jiang D. Digital twin system for manufacturing processes based on a multi-layer knowledge graph model. Sci Rep. 2025;15. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41598-024-85053-0\u003c/span\u003e\u003cspan address=\"10.1038/s41598-024-85053-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSu C, Tang X, Han Y, Wang T, Jiang D. Cognitive digital twin in manufacturing process: integrating the knowledge graph for enhanced human-centric Industry 5.0. Int J Prod Res. 2024;1\u0026ndash;22. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/00207543.2024.2435583\u003c/span\u003e\u003cspan address=\"10.1080/00207543.2024.2435583\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSu C, Han Y, Tang X, Jiang Q, Wang T, He Q. Knowledge-based digital twin system: Using a knowlege-driven approach for manufacturing process modeling. Computers Ind Volumes. 2024. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e104101.https://doi.org/10.1016/j.compind.2024.104101\u003c/span\u003e\u003cspan address=\"104101.10.1016/j.compind.2024.104101\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. 159\u0026ndash;160,2024,.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKarabulut E, Pileggi S, Groth P, Degeler V, ArXiv. abs/2308.15168.\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.future.2023.12.013\u003c/span\u003e\u003cspan address=\"10.1016/j.future.2023.12.013\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAkroyd J, Mosbach S, Bhave A, Kraft M. Universal Digital Twin - A Dynamic Knowledge Graph. Data-Centric Eng. 2021. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1017/dce.2021.10\u003c/span\u003e\u003cspan address=\"10.1017/dce.2021.10\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. 2.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLim K, Yosal T, Chen C, Zheng P, Wang L, Xu X. Graph-enabled cognitive digital twins for causal inference in maintenance processes. Int J Prod Res. 2023;62:4717\u0026ndash;34. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/00207543.2023.2274335\u003c/span\u003e\u003cspan address=\"10.1080/00207543.2023.2274335\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFeng J, Tang H, Zhou S, Cai Y, Zhang J. Cognitive Digital Twins of the natural environment: Framework and application. Eng Appl Artif Intell. 2025;139:109587. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.engappai.2024.109587\u003c/span\u003e\u003cspan address=\"10.1016/j.engappai.2024.109587\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStavropoulou G, Tsitseklis K, Mavraidi L, Chang K, Zafeiropoulos A, Karyotis V, Papavassiliou S. Digital Twin Meets Knowledge Graph for Intelligent Manufacturing Processes. Sensors. 2024;24. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/s24082618\u003c/span\u003e\u003cspan address=\"10.3390/s24082618\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRamonell C, Chac\u0026oacute;n R, Posada H. Knowledge graph-based data integration system for digital twins of built assets. Autom Constr. 2023. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.autcon.2023.105109\u003c/span\u003e\u003cspan address=\"10.1016/j.autcon.2023.105109\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJia M, Hu J, Liu Y, Gao Z, Yao Y, Industrial. \u0026amp; Engineering Chemistry Research.\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1021/acs.iecr.2c03628\u003c/span\u003e\u003cspan address=\"10.1021/acs.iecr.2c03628\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHofmeister M, Bai J, Brownbridge G, Mosbach S, Lee K, Farazi F, Hillman M, Agarwal M, Ganguly S, Akroyd J, Kraft M. Semantic agent framework for automated flood assessment using dynamic knowledge graphs. Data-Centric Eng. 2024;5. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1017/dce.2024.11\u003c/span\u003e\u003cspan address=\"10.1017/dce.2024.11\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMortlock T, Muthirayan D, Yu S, Khargonekar P, Faruque M. Graph Learning for Cognitive Digital Twins in Manufacturing Systems. IEEE Trans Emerg Top Comput. 2021;10:34\u0026ndash;45. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/TETC.2021.3132251\u003c/span\u003e\u003cspan address=\"10.1109/TETC.2021.3132251\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVillalonga A, Negri E, Biscardo G, Casta\u0026ntilde;o F, Haber R, Fumagalli L, Macchi M. A decision-making framework for dynamic scheduling of cyber-physical production systems based on digital twins. Annu Rev Control. 2021;51:357\u0026ndash;73. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.arcontrol.2021.04.008\u003c/span\u003e\u003cspan address=\"10.1016/j.arcontrol.2021.04.008\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChang L, Zhang L, Fu C, Chen Y. Transparent Digital Twin for Output Control Using Belief Rule Base. IEEE Trans Cybernetics. 2021;52:10364\u0026ndash;78. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/TCYB.2021.3063285\u003c/span\u003e\u003cspan address=\"10.1109/TCYB.2021.3063285\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang Y, Dhouib S, Medinacelli L, Malenfant J. (2023). Semantic Interoperability of Digital Twins: Ontology-based Capability Checking in AAS Modeling Framework. 2023 IEEE 6th International Conference on Industrial Cyber-Physical Systems (ICPS), 1\u0026ndash;8. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/ICPS58381.2023.10128003\u003c/span\u003e\u003cspan address=\"10.1109/ICPS58381.2023.10128003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eG\u0026ouml;ppert A, Grahn L, Rachner J, Grunert D, Hort S, Schmitt R. Pipeline for ontology-based modeling and automated deployment of digital twins for planning and control of manufacturing systems. J Intell Manuf. 2021;34:2133\u0026ndash;52. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10845-021-01860-6\u003c/span\u003e\u003cspan address=\"10.1007/s10845-021-01860-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDittler D, Frank P, Hildebrandt G, Peterson L, Jazdi N, Weyrich M, ArXiv. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003eabs/2507.03553.https://doi.org/10.48550/arXiv.2507.03553\u003c/span\u003e\u003cspan address=\"abs/2507.03553.10.48550/arXiv.2507.03553\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBooyse W, Wilke D, Heyns S. Deep digital twins for detection, diagnostics and prognostics. Mech Syst Signal Process. 2020;140:106612. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.ymssp.2019.106612\u003c/span\u003e\u003cspan address=\"10.1016/j.ymssp.2019.106612\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBevilacqua M, Bottani E, Ciarapica F, Costantino F, Di Donato L, Ferraro A, Mazzuto G, Monteri\u0026ugrave; A, Nardini G, Ortenzi M, Paroncini M, Pirozzi M, Prist M, Quatrini E, Tronci M, Vignali G. Digital Twin Reference Model Development to Prevent Operators\u0026rsquo; Risk in Process Plants. Sustainability. 2020;12:1088. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/su12031088\u003c/span\u003e\u003cspan address=\"10.3390/su12031088\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSomers R, Douthwaite J, Wagg D, Walkinshaw N, Hierons R. Digital-twin-based testing for cyber-physical systems: A systematic literature review. Inf Softw Technol. 2022;156:107145.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOmrany H, Al-Obaidi K, Husain A, Ghaffarianhoseini A. (2023). Digital Twins in the Construction Industry: A Comprehensive Review of Current Implementations, Enabling Technologies, and Future Directions. Sustainability.\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.infsof.2022.107145\u003c/span\u003e\u003cspan address=\"10.1016/j.infsof.2022.107145\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu T, Zhang X, Sun W, Wang B. Sensors. 2025;25. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/s25071972\u003c/span\u003e\u003cspan address=\"10.3390/s25071972\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Intelligent Operation and Maintenance of Wind Turbines Gearboxes via Digital Twin and Multi-Source Data Fusion.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAbdullahi I, Longo S, Samie M. (2024). Towards a Distributed Digital Twin Framework for Predictive Maintenance in Industrial Internet of Things (IIoT). Sensors (Basel, Switzerland), 24. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/s24082663\u003c/span\u003e\u003cspan address=\"10.3390/s24082663\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVan Dinter R, Tekinerdogan B, Catal C. Reference architecture for digital twin-based predictive maintenance systems. Comput Ind Eng. 2023;177:109099. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.cie.2023.109099\u003c/span\u003e\u003cspan address=\"10.1016/j.cie.2023.109099\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu X, Wu X, He J, Gupta R. (2025). Intelligent Operation and Maintenance and Prediction Model Optimization for Improving Wind Power Generation Efficiency. 2025 7th International Congress on Human-Computer Interaction, Optimization and Robotic Applications (ICHORA), 1\u0026ndash;7. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.48550/arXiv.2506.16095\u003c/span\u003e\u003cspan address=\"10.48550/arXiv.2506.16095\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLadj A, Wang Z, Meski O, Belkadi F, Ritou M, Da Cunha C. A knowledge-based Digital Shadow for machining industry in a Digital Twin perspective. J Manuf Syst. 2020. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jmsy.2020.07.018\u003c/span\u003e\u003cspan address=\"10.1016/j.jmsy.2020.07.018\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZheng P, Xia L, Li C, Li X, Liu B. Towards Self-X cognitive manufacturing network: An industrial knowledge graph-based multi-agent reinforcement learning approach. J Manuf Syst. 2021;61:16\u0026ndash;26. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jmsy.2021.08.002\u003c/span\u003e\u003cspan address=\"10.1016/j.jmsy.2021.08.002\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Digital Twins, Knowledge Graphs, Q-Layer, four fusion patterns","lastPublishedDoi":"10.21203/rs.3.rs-8401810/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8401810/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eIn modern industrial systems, Digital Twins (DT) focus on mapping physical states, simulating conditions, and predicting operations. Knowledge Graphs (KG) specialize in structured knowledge representation, cross-entity association, and logical inference. Despite their complementarity, the separation between the data and semantic layers leads to challenges such as limited model reuse, weak semantic interpretation, and a lack of closed-loop control. This paper presents a four-layer DT\u0026ndash;KG fusion architecture (Q-Layer) that defines collaborative mechanisms from entity mapping through cognitive decision-making and distills four fusion patterns suited to diverse industrial scenarios. Q-Layer addresses the problem of heterogeneous data silos by establishing a closed loop of perception, semantics, inference, and decision, significantly enhancing system intelligence and adaptability. This framework offers a scalable pathway and theoretical foundation for advancing industrial intelligence.\u003c/p\u003e","manuscriptTitle":"A Digital Twin and Knowledge Graph Fusion Framework for Industrial Intelligence","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-30 12:39:05","doi":"10.21203/rs.3.rs-8401810/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"d06c8acf-a66b-4062-aa69-d6a59b76d7d9","owner":[],"postedDate":"December 30th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-01-14T13:09:36+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-30 12:39:05","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8401810","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8401810","identity":"rs-8401810","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}