Command centers of modern power transmission and distribution systems ingest massive heterogeneous data - structured measurements e.g., Supervisory Control and Data Acquisition ( SCADA), topological models Geographic Information System (GIS), and unstructured operational logs -creating a critical challenge for timely, accurate, and explainable decision support. Existing single -tool or siloed approaches fail to jointly interpret natural language querie s over hybrid data while ensuring reliability and traceability. This study introduces Hybrid Data Agent (HAL), a centralized multi - agent Large Language Model (LLM) system designed for natural language hybrid data querying and decision support in power grid command environments. HAL employs an LLM -based central coordinator to decompose user queries into subtasks and dispatch them to specialized expert agents -such as Structured Query Language (SQL) execution, semantic document retrieval, time -series anomaly d etection, and causality reasoning -then fuses their outputs into coherent, verified responses. The system is evaluated on a high -fidelity simulated benchmark reflecting realistic grid operational scenarios, including electrical - physical events (e.g., curren t surges and protection switching). Comparative experiments against baselines (generic LLM agents and traditional pipelines) show that HAL substantially improves query success rate, answer accuracy, and robustness, while maintaining explainability via stru ctured “think -act -observe” execution logs. Contributions include: 1) a distributed yet centrally coordinated architecture combining LLM planning with deterministic domain tools; 2) a hybrid data fusion mechanism aligning structured and unstructured evidenc e for natural language interfaces; 3) an explainable decision support engine tailored to power grid command centers. HAL reduces cognitive load on operators, accelerates fault diagnosis, and strengthens real -time operational intelligence, making it suitabl e for deployment in intelligent information systems for critical infrastructure.

[1] Al Mtawa Y., Haque A., and Halabi T., “A Review and Taxonomy on Fault Analysis in Transmission Power Systems,” Computation , vol. 10, no. 9, pp. 1 -19, 2022. DOI: 10.3390/computation 10090144

[2] Avcı E., “Hybrid Artificial Intelligence Techniques for Enhanced Electricity Outage Prediction and Management in Distribution Networks,” Turk J Electr Power Energy Syst , vol. 4, no. 2, pp. 63 -73, 2024. DOI:10.5152/tepes.2024.24008

[3] Cen J., Liu J., Li Z., and Wang J., “SQLFixAgent: Towards Semantic -Accurate Text -to-SQL Parsing via Consistency -Enhanced Multi -Agent Collaboration,” in Proceedings of the AAAI Conference on Artificial Intelligence , Philadelphia, pp. 49 -57, 2025. DOI: 10.1609/aaai.v39i1.31 979

[4] Cheng Y., Zhao H., Zhou X., Zhao J., and et al., “A Large Language Model for Advanced Power Dispatch,” Scientific Reports , vol. 15, no. 1, Article 8925, 2025. DOI: 10.1038/s41598 -025 - 91940 -x

[5] Garg M., Namboodiri N., Joshi K., Krishna K., and Vora D., “O ptimizing Multimodal RAG Systems for Multilingual Product Support,” The International Arab Journal of Information Technology , vol. 23, no. 3, pp. 505 -520, 2026. DOI: 10.34028/iajit/23/3/9

[6] Gholamzadeh Khoee A., Wang S., Yu Y., Feldt R., and Parthasarathy D. , “GateLens: A Reasoning -Enhanced LLM Agent for Automotive Software Release Analytics,” arXiv Preprint , arXiv:2503.21735v4, 2025. DOI: 10.48550/arXiv.2503.21735

[7] Ji Z., Lee N., Frieske R., Yu T., and et al., “Survey of Hallucination in Natural Language Gene ration,” ACM Computing Surveys , vol. 55, no. 12, pp. 1 -38, 2023. DOI: 10.1145/3571730

[8] Ko K., Nyein T., Oo K., Oo T., and Zin T., “Retrieval Augmented Generation for Document Query Automation Using Open Source LLMs,” in Proceedings of the 5th International Conference on Advanced Information Technologies , Yangon, pp. 1 -6, 2024. DOI: 10.1109/ICAIT65209.2024.10754919

[9] Lee U., Kim Y., Lee S., Park J., and et al., “Can We Use GPT -4 as a Mathematics Evaluator in Education?: Exploring the Efficacy and Limitation of LLM -Based Automatic Assessment System for Open -Ended Mathematics Question,” International Journal of Artificial Intelligence in Education , vol. 35, no. 3, pp. 1 -37, 2024. DOI: 10.1007/s40593 -024 -00448 -4

[10] Li Y., Wang C., Li J., Liu J., and Chen Y., “Big Data -Driven Smart Energy Management: From Applications to Challenges,” Journal of Cleaner Production , vol. 317, no. 128369, pp. 215 -225, 2021. https://doi.org/10.1016/j.rser.2015.11.050

[11] Mahmud S., Nakamura M., and Zilberstein S., “Maple: A Framework for Active Preference Learning Guided by Large Language Models,” in Proceedings of the AAAI Conference on Artificial Intelligence , Philadelphia, pp. 27518 -27528, 2025. https://doi.org/10.1609/aaai.v39i26.34964

[12] Piccialli F., Chiaro D., Sarwar S., Cerciello D., and et al., “AgentAI: A Comprehensive Survey on Autonomous Agents in Distributed AI for Industry 4.0,” Expert Systems with Applications , vol. 291, no. C, pp. 1 -18, 2025. DOI: 10.1016/j.eswa.2025.128404

[13] Qiao J., Zhou A., Peng L., Zhu L., and et al., “Research and Implementation of Hybrid Storage Method for Multi -Source Heterogeneous Data of Electric Distribution Network,” in Proceedings of the 3D Imaging -Multidimensional Signal Processing and Deep Learning: Images, Augmented Reality and Information Technologies , Springer, pp. 31 -40, 2023. DOI: 10.1007/978 -981 -99 -1230 -8_3

[14] Sequeda J., Allemang D., and Jacob B., “A Benchmark to Understand the Role of Knowledge Graphs on Large Language Model’s Accuracy for Question Answering on Enterprise SQL Databases,” in Proceedings of the 7 th Joint Workshop on Graph Data Management Experiences and Systems and Network Data Analytics , New York, pp. 1 -12, 2024. https://doi.org/10.1145/3661304.3661901

[15] Singh A., Shetty A., Ehtesham A., Kumar S., and Khoei T., “A Survey of Large Language Model - Based Generative AI for Text -to-SQL: Benchmarks, Applications, Use Cases, and Challenges,” in Proceedings of the IEEE 15 th Annual Computing and Communication Workshop and Conference , Las Vegas, pp. 15 -21, 2025. DOI:10.1109/CCWC62904.2025.10903689

[16] Sur i S., Das S., Singi K., Dey K., and et al., “Software Engineering Using Autonomous Agents: Are We There Yet?,” in Proceedings of the 38 th IEEE/ACM International Conference on Automated Software Engineering , Echternach, pp. 1855 -1857, 2023. DOI: 10.1109/ASE 56229.2023.00174

[17] Wan Y., Chen Z., Liu Y., Chen C., and Packianather M., “Empowering LLMs by Hybrid Retrieval -Augmented Generation for Domain - Centric Q and A in Smart Manufacturing,” Advanced Engineering Informatics , vol. 65, no. PB, pp. 1 -16, 2025. https:/ /doi.org/10.1016/j.aei.2025.103212

[18] Wang J. and Zhao C., “Robust Control Performance Monitoring for Varying - Dimensional Time -Series Data Based on SCADA Systems,” IEEE Transactions on Instrumentation and Measurement , vol. 71, pp. 1 - 14, 2022. DOI: 10.1109/TIM .2022.3177217

[19] Wang L., Ma C., Feng X., Zhang Z., and et al., “A Survey on Large Language Model Based Autonomous Agents,” arXiv Preprint , arXiv:2308.11432v7, pp. 1 -42, 2023. https://doi.org/10.1007/s11704 -024 -40231 -1

[20] Wu S., Wei W., Zhang M., Chen Z., and et al., “Generative Retrieval as Multi -Vector Dense Retrieval,” in Proceedings of the 47 th International ACM SIGIR Conference on Research and Development in Information Retrieval , New York, pp. 1828 -1838, 2024. DOI: 10.1145/3626772.3657697

[21] Zhang B. and Yang Y., “MediaHG: Rethinking Eye -Catchy Features in Social Media Headline Generation,” in Proceedings of the Conference on Empirical Methods in Natural Language Processing , Singapore, pp. 5766 -5777, 2023. DOI: 10.18653/v1/2023.emnlp -main.352