Energy Informatics最新文献

{"title":"MDJA-Trans: a multivariate data joint analysis transformer architecture with dynamic gating for load-aware low-voltage AC Arc fault detection","authors":"Jing Li,&nbsp;Nawaraj Kumar Mahato,&nbsp;Yubin Guo,&nbsp;Junfeng Yang","doi":"10.1186/s42162-026-00663-4","DOIUrl":"10.1186/s42162-026-00663-4","url":null,"abstract":"<div>\u0000 \u0000 <p>To address the challenges of multivariate signal coupling, noise interference, and load adaptability in arc fault detection within low-voltage distribution systems, this paper introduces a novel arc fault detection architecture, termed Multivariate Data Joint Analysis with Transformer (MDJA-Trans) architecture. The architecture employs Noise-Suppressing Time Convolutional Learning (NS-TCL) to dynamically mitigate noise and extract multiscale arc features through frequency-domain masking. Load-Aware Cross-Variate Attention Fusion (CVAF) facilitates dynamic alignment of voltage and current features, while Dynamic Gating Decision (DGD) integrates electrical features with load attribute embeddings for robust arc fault classification. This approach enhances physical interpretability and load generalization. Extensive experiments demonstrate superior noise immunity and cross-load adaptability, achieving an arc fault detection accuracy and F1-Score of 98.55%. The proposed solution will provide technical support for improving household safety and reducing electrical fire risks.</p>\u0000 </div>","PeriodicalId":538,"journal":{"name":"Energy Informatics","volume":"9 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2026-03-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1186/s42162-026-00663-4.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147829913","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A multi-sensor informatics workflow for appliance-level energy use, comfort, and mixed-mode cooling in a composite-climate home","authors":"Sahil Chilana,&nbsp;Aviruch Bhatia,&nbsp;Vishal Garg,&nbsp;Jyotirmay Mathur","doi":"10.1186/s42162-026-00662-5","DOIUrl":"10.1186/s42162-026-00662-5","url":null,"abstract":"<div>\u0000 \u0000 <p>This study presents a year-long, multi-sensor residential monitoring dataset from a composite-climate home in Mohali, India, integrating appliance-level electricity use, indoor temperature–humidity measurements, and outdoor weather observations into an hourly residential energy informatics workflow. Nine appliances were monitored at 5-minute resolution and processed through a reproducible pipeline involving timestamp harmonization, anomaly handling, multi-source alignment, and feature extraction, yielding an analytics-ready dataset spanning 8,736 hourly observations (July 2024–June 2025). The monitored appliances consumed 845.25 kWh (58% of total household electricity use), with the refrigerator (21.95%) and air conditioner (cooling + heating; 19.07%) as the dominant monitored end uses. Psychrometric mapping quantified mixed-mode operation, distinguishing operating regions for Air Conditioner (AC) cooling (n=460 hours), evaporative cooling (n=172 hours), and AC heating (n=45 hours). Weather sensitivity analysis showed strong temperature dependence in baseline and seasonal loads (effect of outside temperature on refrigerator r=0.793; effect of outside temperature on geyser r=-0.694; slope -0.048 kWh/<span>(^{circ })</span>C), demonstrating the value of fusing weather variables into appliance analytics. Indoor comfort evaluation using the India Model for Adaptive Comfort (IMAC) indicated that 80.8% of summer hours and 51.7% of monsoon hours with valid measurements fell within adaptive comfort limits, with winter interpretation constrained by modeling assumptions and sensor availability. The primary contribution is a transparent, transferable workflow that converts heterogeneous residential sensing streams into a unified analytics dataset suitable for season-responsive appliance modeling, comfort-aware interpretation, and future multi-home deployments.</p>\u0000 </div>","PeriodicalId":538,"journal":{"name":"Energy Informatics","volume":"9 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2026-03-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1186/s42162-026-00662-5.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147830027","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Towards carbon-aware AI: a systematic prisma review and taxonomy of green architectures, hardware life-cycle, and energy-efficient algorithms","authors":"Raghavendra M. Devadas,&nbsp;Sowmya T","doi":"10.1186/s42162-026-00651-8","DOIUrl":"10.1186/s42162-026-00651-8","url":null,"abstract":"<div>\u0000 \u0000 <p>The increasing computational requirements in modern Artificial Intelligence (AI) development have raised the stakes when it comes to the environmental sustainability of machine learning use and application. The topic that has received relatively little treatment in previous research is the carbon footprint of AI systems, but it is not studied continuously across algorithms, hardware, life cycle, and use. This work takes stock of the state-of-the-art to provide a comprehensive review of carbon-aware AI from the bottom up for the entire gamut of computation. Consistent with PRISMA principles, by way of systematic search across prominent academic databases and repositories (2018–2025), we identified 784 unique citations and extracted 62 studies that satisfied pre-established inclusion criteria. These studies were organized by four domains: algorithms for energy (20), hardware and accelerators (15), Life-cycle assessment (LCA) (9), and operation under deployment (11). Three major conclusions follow from the synthesis. First, algorithmic efficiency – including pruning, quantization, and sparsity might reduce computational burden to meet carbon goals; however, they only achieve concrete carbon reduction when accounted for in hardware and data centre setups. Second, life-cycle analyses show that while operational energy continues to account for most emissions during large-scale training, the embodied carbon from semiconductor fabrication plays an increasingly important role in fleets with a lot of equipment or frequent refreshes. Third, deployment decisions such as data center location, carbon-aware scheduling, and cloud–edge workloads placement bring much more variance on real emissions compared to what can be achieved at model-level optimisation. Between sectors, inconsistencies in methodology – notably for carbon reporting, system boundaries, and energy telemetry – hinder reproducibility and comparison of findings across studies. To overcome these limitations, this review suggests a research agenda with respect to standardized carbon accounting, hardware–software co-optimization, better available embodied-emission data, and the inclusion of carbon in decision-making for AutoML and scheduling systems. This comprehensive integration will serve as a stepping-stone towards the development of sustainable AI across academia and industry.</p>\u0000 </div>","PeriodicalId":538,"journal":{"name":"Energy Informatics","volume":"9 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2026-03-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1186/s42162-026-00651-8.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147561097","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"ESM-Regio: simulation and optimization of regional energy systems in carbon-neutral scenarios","authors":"Kevin-Martin Aigner,&nbsp;Peter Bazan,&nbsp;Sebastian Bottler,&nbsp;Robert Burlacu,&nbsp;Berenike Bölting,&nbsp;Nora Elhaus,&nbsp;Jürgen Karl,&nbsp;Frauke Liers,&nbsp;Natalia Luna-Jaspe,&nbsp;Klaus Markolf,&nbsp;Erich Maurer,&nbsp;Marco Pruckner,&nbsp;Daniel Scharrer,&nbsp;Leo Strobel,&nbsp;Christian Weindl,&nbsp;Reinhard German","doi":"10.1186/s42162-026-00648-3","DOIUrl":"10.1186/s42162-026-00648-3","url":null,"abstract":"<div><p>This work presents a high-resolution, sector-coupled energy system modeling framework developed within the research project ESM-Regio, designed to analyze regional energy transitions and the sector coupling potential in the context of Germany’s 2045 climate targets. In contrast to existing regional energy system models, our framework explicitly models medium-voltage distribution grids and incorporates asset aging into a sector-coupled techno-economic optimization. The framework operates on the medium-voltage level at a 15-minute temporal resolution and is designed to analyze regions ranging city to county (NUTS-3 level). The electricity, heat, transport, and gas sectors are integrated into the framework using a component-based simulation architecture. All sectors are represented with sectoral submodels which are iteratively coupled through alternating simulation and mixed-integer optimization stages to minimize operational system costs while accounting for technical grid constraints and asset aging. The methodology is applied to the distribution grid area of the regional utility Stadtwerke Bayreuth, for three representative weeks in each of the years 2019, 2030, and 2045. The results highlight the increasing importance of coordinated flexibility—such as building thermal inertia, battery storage, and electric vehicles—in mitigating grid stress under increased electrification. Across the investigated scenarios, coordinated flexibility usage reduces energy procurement costs by up to 22% and significantly lowers line and transformer loading. However, we find that while intelligent flexibility usage can reduce the need for grid expansion, it can not completely eliminate it in the carbon-neutral 2045 scenarios. The presented framework offers a scalable, data-driven approach to assess the impacts of sector coupling and to support infrastructure and regional energy transition planning aligned with national decarbonization goals.</p></div>","PeriodicalId":538,"journal":{"name":"Energy Informatics","volume":"9 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2026-03-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1186/s42162-026-00648-3.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147642921","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Intelligent detection of topological relationships of substation electrical main wiring diagram based on CIM/SVG","authors":"Qiang Sun,&nbsp;Zhiqing Song,&nbsp;Zheng Lv,&nbsp;Qi Xiao,&nbsp;Yanqian Lu","doi":"10.1186/s42162-026-00657-2","DOIUrl":"10.1186/s42162-026-00657-2","url":null,"abstract":"<div>\u0000 \u0000 <p>Due to the complexity of substation systems and the diversity of data sources, potential inaccuracies or inconsistencies in data may compromise the accuracy of intelligent detection. To address this issue, an intelligent detection method for topological relationships in substation electrical main wiring diagrams is developed based on the integration of the Common Information Model (CIM) and Scalable Vector Graphics (SVG). First, main wiring diagrams are classified into three structural types: chain, ring, and network. Second, substation conductive equipment is categorized, and the primary node–node correlation matrix is extracted. Third, redundant regions in the wiring diagram images are segmented. By integrating CIM and SVG, the electrical main wiring diagram is automatically generated. Subsequently, component names and coordinates are obtained using the Faster R-CNN deep learning model. A key contribution of this work is the introduction of a dynamic topology prediction module based on a Temporal Graph Convolutional Network (TGCN), which enables real-time adaptation to changes in substation operation modes, thereby enhancing system robustness and operational stability. Finally, topological relationships are intelligently detected using graph theory and adjacency matrices. Experimental results show: correlation matrix extraction time &lt; 20 ms, accuracy = 85.02%, F1-score = 85.42%, average precision (AP) of components &gt; 0.8, and a 91.59% improvement in detection accuracy.</p>\u0000 </div>","PeriodicalId":538,"journal":{"name":"Energy Informatics","volume":"9 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2026-03-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1186/s42162-026-00657-2.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147796468","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Control strategies for district heating and cooling systems: a comprehensive review across production, distribution, and end-user levels","authors":"Nima Monghasemi,&nbsp;Amir Vadiee,&nbsp;Stavros Vouros,&nbsp;Konstantinos Kyprianidis","doi":"10.1186/s42162-026-00650-9","DOIUrl":"10.1186/s42162-026-00650-9","url":null,"abstract":"<div>\u0000 \u0000 <p>The evolution of district heating and cooling systems into sophisticated energy networks is essential for global decarbonization. However, a fundamental tension exists between rapid innovation in advanced control algorithms and the slow replacement cycle of physical infrastructure, making intelligent system-wide control the primary enabler of network modernization. Despite this critical role, the existing literature remains fragmented and lacks a comprehensive synthesis of control strategies across the production, distribution, and end-user levels. This analysis confirmed a definitive shift from isolated, single-level control to holistic frameworks that unlock system-wide flexibility. This review establishes that successful implementation requires addressing distinct objectives at each operational level, from multisource management in production to occupant-centric control at the end-user level. A critical finding is the credibility gap between the demonstrated potential of advanced control and its limited practical application. This disparity is rooted in systemic challenges, including intensive modeling requirements, computational scalability limits, and unresolved human-in-the-loop problems. To bridge this gap, this review presents a structured framework that synthesizes the current state of district heating and cooling control and proposes a forward-looking research roadmap. This roadmap prioritizes the development of hybrid intelligent controllers that integrate learning-based methods with model predictive control, creating persistent digital twins for human-centric applications, and designing secure and decentralized coordination architectures for next-generation thermal networks.</p>\u0000 </div>","PeriodicalId":538,"journal":{"name":"Energy Informatics","volume":"9 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2026-03-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1186/s42162-026-00650-9.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147642910","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Evaluation of energy demands in future scenarios for a residential district via co-simulation of heat, electricity, and mobility: the case of the district “Am Ölper Berge”","authors":"Fernando Penaherrera,&nbsp;Ihsan Ünal,&nbsp;Lars Kühl,&nbsp;Astrid Nieße","doi":"10.1186/s42162-026-00658-1","DOIUrl":"10.1186/s42162-026-00658-1","url":null,"abstract":"<div><h3>Purpose</h3><p>This study examines the thermal, electrical, and mobility energy demand and supply of the residential district “Am Ölper Berge\" in Lower Saxony, and evaluates potential developments of the district energy supply system and the trade-offs associated with its modernization.</p><h3>Methods</h3><p>A reference scenario for the year 2020 serves as the baseline for evaluating future development points (2030, 2040, 2050), including increasing electrification of thermal energy supply, improvements in building energy efficiency, and the integration of renewable energy sources for local generation. The scenarios include step-by-step measures such as facade insulation, implementation of photovoltaic systems, district heating supply, implementation of heat pumps, and the integration of charging infrastructure for electric vehicles. This is achieved using several control models integrated into the co-simulation platform mosaik.</p><h3>Results</h3><p>In the 2050 scenario, heat supply is provided via heat pumps and a low-temperature district heating network, supplemented by neighborhood energy storage and extensive use of photovoltaic systems. Four extended scenarios examine centralized and decentralized control strategies that monitor grid voltage across the district and optimize electric vehicle charging. The analysis includes charging strategies such as maximum charging power, night-time charging, forecast-based charging, and solar-optimized charging. The results show significant differences in external energy demands and CO<span>(_2)</span> emissions, with solar-optimized charging of electrical vehicles combined with areal use of power flexibilities for grid control offering the greatest environmental and operational benefits.</p><h3>Conclusion</h3><p>The study provides key insights into the interactions between building energy demands, decentralized generation, and grid operation, thereby supporting the planning of energy supply for urban districts in alignment with emission reduction targets.</p></div>","PeriodicalId":538,"journal":{"name":"Energy Informatics","volume":"9 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2026-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1186/s42162-026-00658-1.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147643016","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Explainable AI-driven energy forecasting: A DGMR-based feature extraction and EGST-Net prediction framework for transparent decision-making","authors":"Swaroopa Rani B,&nbsp; Dr. Chandrashekar Jatoth,&nbsp;Dr. Sonti Venu","doi":"10.1186/s42162-026-00656-3","DOIUrl":"10.1186/s42162-026-00656-3","url":null,"abstract":"<div>\u0000 \u0000 <p>Transparency in Decision-Making the modern energy management systems, Transparency in Decision-Making requires proper and understandable forecasting models capable of controlling and documenting the intricate dynamics of consumption when implemented within the smart grid scenario. However, the existing energy prediction techniques, including traditional statistical models and independent deep learning techniques, are often incapable of learning a nonlinear spatiotemporal relationship, and offer poor interpretability, resulting in a reduction of reliability to real-life use. This paper proposes an Explainable AI-based energy forecasting model to address these limitations, which will be a combination of the Dynamic Gated Memory Refinement (DGMR) and a hybrid EGST-Net framework comprising Convolutional Long Short-Term Memory (ConvLSTM) and Transformer-based attention models. The suggested method is novel in a sense that it is adaptably capable of refining feature of its aim that chooses selectively, in informative temporal-spatial patterns and eliminates noise, and an attention-driven elucidation section that enhances intelligibility in the determination to make forecasts.The framework conducts the multi-stage processing of the data (data preprocessing, DGMR-based feature extraction, spatiotemporal learning with the help of EGST-Net, and attention-based interpretability analysis) and was launched on the Python environment with deep learning libraries to provide effective training and testing of the model. The experimental findings showed better performance than the traditional baseline models with a prediction accuracy of 96% and little forecasting error scores. The suggested system will appeal to energy providers, smart grid operators, policymakers, and other industrial stakeholders as it can facilitate resource allocation and efficient demand prediction, as well as make the decisions made by AI interpretable. All in all, the framework is part of the achievement of intelligent, transparent, and scalable energy management solutions to the development of next generation smart energy infrastructures.</p>\u0000 </div>","PeriodicalId":538,"journal":{"name":"Energy Informatics","volume":"9 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2026-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1186/s42162-026-00656-3.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147737669","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Measurement of green development energy level and analysis of obstacle factors in the Yangtze River Delta urban agglomeration based on the improved WRSR model and obstacle degree model","authors":"Hanjie Xiao,&nbsp;Hui Jiang,&nbsp;Qianhui Bao,&nbsp;Liang Wu","doi":"10.1186/s42162-026-00653-6","DOIUrl":"10.1186/s42162-026-00653-6","url":null,"abstract":"<div>\u0000 \u0000 <p>Accurately measuring the green development energy level of the Yangtze River Delta urban agglomeration and identifying its key drivers are crucial for promoting regional sustainable development, yet they remain complex challenges. This study first establishes a comprehensive evaluation index system covering four dimensions: green economy, green innovation, green support, and green openness. Combined weights are determined by integrating the G1 method and entropy method through game theory, and an optimized Non-integer Weighted Rank-Sum Ratio (WRSR) evaluation model is proposed. Subsequently, using the improved WRSR model and obstacle degree model, the green development energy levels of 27 representative cities in the region from 2006 to 2021 are measured and analyzed. The results indicate that: (1) The overall green development energy level of the agglomeration increased slowly from 0.275 in 2006 to 0.300 in 2021, with significant disparities among cities. (2) Cities are classified into four tiers, with Shanghai and Nanjing as higher-level cities, Wuxi, Changzhou, and Suzhou as high-level cities, Chuzhou, Xuancheng, and Anqing as low-level cities, and Chizhou as a lower-level city, revealing a core-periphery spatial pattern. (3) Green openness and green innovation are key influencing factors, with obstacles mainly arising from indicators such as foreign trade dependence, actual utilized foreign direct investment, and the number of full-time teachers in higher education institutions. This research provides a methodological reference for evaluating urban green development and offers targeted policy insights to support coordinated green transition across city tiers.</p>\u0000 </div>","PeriodicalId":538,"journal":{"name":"Energy Informatics","volume":"9 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2026-03-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1186/s42162-026-00653-6.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147737447","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Identification model of distribution equipment insulation aging enhancement based on SCADA knowledge graph","authors":"Shuai Zhang,&nbsp;Wei Zhang,&nbsp;Song Wang,&nbsp;Lianwei Bao,&nbsp;Zhou Yu","doi":"10.1186/s42162-026-00639-4","DOIUrl":"10.1186/s42162-026-00639-4","url":null,"abstract":"<div><p>With the continuous advancement of scientific and technological integration in power facilities, higher requirements have been raised for identifying the insulation aging state of distribution equipment. At present, Supervisory Control and Data Acquisition (SCADA) systems face bottlenecks due to the limited information dimensions of single-sensor data and the heavy computational burden of complex models, which restrict their deployment and application in practical scenarios. To address these challenges, a multimodal data fusion framework is introduce and collaborative analysis and feature extraction are performed on monitoring signals from different physical characteristics. Furthermore, a lightweight Knowledge Graph-Enhanced Dynamic Graph Neural Network (KGE-DGNN) is innovatively proposed by integrating an adaptive feature weighting module. This model can autonomously enhance the contribution of key modalities while maintaining efficient computational logic, significantly reducing resource consumption and improving the overall performance of insulation aging identification. Experimental results demonstrate that the recognition accuracy reaches 98.5%, which is 8% higher than that of the baseline method. The computational efficiency achieves an average single recognition time of 120 ms. Moreover, the peak memory occupancy remains below 350 MB, which fully validates its application potential in real-time diagnostic scenarios and considerably improves the balance between accuracy and efficiency. Thus, the proposed method provides a novel and reliable intelligent diagnosis tool for early fault warning in distribution equipment. Its technical approach holds great value in promoting the development of condition-based maintenance toward precision and intelligence.</p></div>","PeriodicalId":538,"journal":{"name":"Energy Informatics","volume":"9 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2026-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1186/s42162-026-00639-4.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147642674","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}