The global transition to clean energy has intensified demand for intelligent wind power systems, yet operational challenges persist due to fragmented knowledge and limited integration across business processes and maintenance management. While large language models (LLMs) exhibit strong reasoning and orchestration capabilities, their application in wind power systems remains constrained by inefficient retrieval-augmented generation (RAG) over heterogeneous knowledge bases, insufficient multimodal data integration, and limited edge resources. To address these challenges, we propose a MetaGPT-based LLM agent framework that dynamically integrates tools, lightweight models, and domain-specific databases to enable seamless knowledge orchestration. The framework introduces mix-augmented generation (MAG), which combines cache-augmented generation (CAG) with an optimized RAG pipeline to enhance retrieval efficiency and semantic alignment, outperforming both RAG and CAG through adaptive knowledge integration and dynamic query routing. In addition, the framework employs an activation-aware weight quantization strategy to reduce model size and memory usage while maintaining minimal accuracy loss. Evaluated on the EleQA wind energy dataset, the proposed framework demonstrates coordinated reasoning and task execution capabilities within the multi-agent system, while significantly reducing retrieval latency and accelerating inference. These results highlight its effectiveness and efficiency for real-time edge deployment in wind power system operations.
| [1] |
Hassan Q., Algburi S., Sameen A.Z., Salman H.M., Jaszczur M.. A review of hybrid renewable energy systems: solar and wind-powered solutions: challenges, opportunities, and policy implications. Results Eng., 2023, 20. 101621
|
| [2] |
Noussis A., O’Neil R., Saif A., Khatab A., Diallo C.. A hybrid bi-lstm model for data-driven maintenance planning. Auton. Intell. Syst., 2025, 5113.
|
| [3] |
He C., Geng H., Rajashekara K., Chandra A.. Analysis and control of frequency stability in low-inertia power systems: a review. IEEE/CAA J. Autom. Sin., 2024, 111223632383.
|
| [4] |
Hong Q., Khan M.A.U., Henderson C., Egea-Àlvarez A., Tzelepis D., Booth C.. Addressing frequency control challenges in future low-inertia power systems: a Great Britain perspective. Engineering, 2021, 7810571063.
|
| [5] |
Xiao A., Yan W., Zhang X., Liu Y., Zhang H., Liu Q.. Multi-domain fusion for cargo uav fault diagnosis knowledge graph construction. Auton. Intell. Syst., 2024, 4110.
|
| [6] |
Bhusal N., Abdelmalak M., Kamruzzaman M., Benidris M.. Power system resilience: current practices, challenges, and future directions. IEEE Access, 2020, 8: 18064-18086.
|
| [7] |
Yu T., Tang J., Yu Q., Li L., Liu Y., Poler R.. Large language models for phm: a review of optimization techniques and applications. Auton. Intell. Syst., 2025, 51114.
|
| [8] |
Peng C., Xia F., Naseriparsa M., Osborne F.. Knowledge graphs: opportunities and challenges. Artif. Intell. Rev., 2023, 56111307113102.
|
| [9] |
Tang Y., Han H., Yu X., Zhao J., Liu G., Wei L.. An intelligent question answering system based on power knowledge graph. 2021, 2021IEEE Power & Energy Society General Meeting (PESGM)0105IEEE
|
| [10] |
Wang F., Zhang X., Tang R., Jiang J.. Wt-fault-llm: knowledge graph enhanced large model for wind-turbine faults. Proc. Comput. Sci., 2025, 266: 385-390.
|
| [11] |
Jin B., Xu X.. China commodity price index (ccpi) forecasting via the neural network. Int. J. Financ. Eng., 2025, 1232550003
|
| [12] |
Li J., Xu X.. Forecasts of coking coal futures price indices through Gaussian process regressions. Miner. Econ., 2025, 381101112.
|
| [13] |
Liu C., Xu X.. Contemporaneous causal orderings among prices of retail properties: evidence from Chinese cities through vector error-correction modeling and directed acyclic graphs. J. Financ. Manag. Prop. Constr., 2025.
|
| [14] |
Xu X., Zhang Y.. Individual time series and composite forecasting of the Chinese stock index. Mach. Learn. Appl., 2021, 5. 100035
|
| [15] |
Chen F., Liu L., Tang B., Chen B., Xiao W., Zhang F.. A novel fusion approach of deep convolution neural network with auto-encoder and its application in planetary gearbox fault diagnosis. Proc. Inst. Mech. Eng., Part O: J. Risk Reliab., 2021, 2351316
|
| [16] |
Jiang L., Zheng C., Li Y.. Rotating machinery fault diagnosis based on transfer learning and an improved convolutional neural network. Meas. Sci. Technol., 2022, 3310105012
|
| [17] |
Xu X.. Corn cash price forecasting. Am. J. Agric. Econ., 2020, 102412971320.
|
| [18] |
W.X. Zhao, K. Zhou, J. Li, T. Tang, X. Wang, Y. Hou, Y. Min, B. Zhang, J. Zhang, Z. Dong, et al., A survey of large language models, 1(2) (2023). arXiv preprint arXiv:2303.18223
|
| [19] |
Xi Z., Chen W., Guo X., He W., Ding Y., Hong B., Zhang M., Wang J., Jin S., Zhou E., et al. The rise and potential of large language model based agents: a survey. Sci. China Inf. Sci., 2025, 682121101
|
| [20] |
Hong S., Zhuge M., Chen J., Zheng X., Cheng Y., Wang J., Zhang C., Wang Z., Yau S.K.S., Lin Z., et al. Metagpt: meta programming for a multi-agent collaborative framework. The Twelfth International Conference on Learning Representations, 2024
|
| [21] |
P. Zhao, H. Zhang, Q. Yu, Z. Wang, Y. Geng, F. Fu, L. Yang, W. Zhang, J. Jiang, B. Cui, Retrieval-augmented generation for ai-generated content: a survey (2024). arXiv preprint arXiv:2402.19473
|
| [22] |
Y. Gao, Y. Xiong, X. Gao, K. Jia, J. Pan, Y. Bi, Y. Dai, J. Sun, H. Wang, H. Wang, Retrieval-augmented generation for large language models: a survey, 2(1) (2023). arXiv preprint arXiv:2312.10997
|
| [23] |
J. Linders, J.M. Tomczak, Knowledge graph-extended retrieval augmented generation for question answering (2025). arXiv preprint arXiv:2504.08893
|
| [24] |
Z. Wang, Z. Wang, L. Le, H.S. Zheng, S. Mishra, V. Perot, Y. Zhang, A. Mattapalli, A. Taly, J. Shang, et al., Speculative rag: enhancing retrieval augmented generation through drafting (2024). arXiv preprint arXiv:2407.08223
|
| [25] |
A. Asai, Z. Wu, Y. Wang, A. Sil, H. Hajishirzi (2024). Self-rag: Learning to retrieve, generate, and critique through self-reflection
|
| [26] |
Z. Guo, L. Xia, Y. Yu, T. Ao, C. Huang, Lightrag: simple and fast retrieval-augmented generation (2024). arXiv preprint arXiv:2410.05779
|
| [27] |
R. Nogueira, K. Cho, Passage re-ranking with bert (2019). arXiv preprint arXiv:1901.04085
|
| [28] |
A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A.N. Gomez, Ł. Kaiser, I. Polosukhin, Attention is all you need. Adv. Neural Inf. Process. Syst. 30 (2017)
|
| [29] |
Chan B.J., Chen C.-T., Cheng J.-H., Huang H.-H.. Don’t do rag: when cache-augmented generation is all you need for knowledge tasks. Companion Proceedings of the ACM on Web Conference 2025, 2025893897.
|
| [30] |
Lin J., Tang J., Tang H., Yang S., Chen W.-M., Wang W.-C., Xiao G., Dang X., Gan C., Han S.. Awq: activation-aware weight quantization for on-device llm compression and acceleration. Proc. Mach. Lear. Syst., 2024, 6: 87-100
|
| [31] |
Wang H., Wei J., Jing H., Song H., Xu B.. Meta-rag: a metadata-driven retrieval-augmented generation framework for the power field. Comput. Eng., 2025, 1110070415
|
| [32] |
A. Yang, B. Yang, B. Zhang, B. Hui, B. Zheng, B. Yu, et al. (2025). Qwen2.5 technical report. arXiv preprint arXiv:2412.15115
|
| [33] |
R. Gong, Y. Yong, S. Gu, Y. Huang, C. Lv, Y. Zhang, X. Liu, D. Tao, Llmc: benchmarking large language model quantization with a versatile compression toolkit (2024). arXiv preprint arXiv:2405.06001
|
| [34] |
T. Zhang, V. Kishore, F. Wu, K.Q. Weinberger, Y. Artzi, Bertscore: evaluating text generation with bert (2019). arXiv preprint arXiv:1904.09675
|
Funding
National Natural Science Foundation of China(62503369)
Shanghai Rising-Star Program (Sailing Special Program)(24YF2749000)
Science and Technology Research Collaboration Project of the Fuzhou Science and Technology Bureau(2025-Y-004)
Rights & permissions
The Author(s)
PDF
