![[2602.21715] Two-Stage Active Distribution Network Voltage Control via LLM-RL Collaboration: A Hybrid Knowledge-Data-Driven Approach](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.21715] Two-Stage Active Distribution Network Voltage Control via LLM-RL Collaboration: A Hybrid Knowledge-Data-Driven Approach
Summary
This article presents a hybrid approach for voltage control in active distribution networks, combining large language models and reinforcement learning to enhance efficiency and performance.
Why It Matters
As distributed photovoltaics become more prevalent, managing voltage levels in active distribution networks is critical. This research offers a novel solution that integrates advanced AI techniques, potentially improving power quality and operational efficiency in real-world scenarios.
Key Takeaways
- The proposed hybrid approach utilizes LLM and RL for effective voltage control.
- A two-stage process enhances both forecasting and real-time adjustments.
- The method improves training efficiency and voltage control performance.
- Incorporates diverse operational data for better decision-making.
- Demonstrated effectiveness through comprehensive comparisons and studies.
Electrical Engineering and Systems Science > Systems and Control arXiv:2602.21715 (eess) [Submitted on 25 Feb 2026] Title:Two-Stage Active Distribution Network Voltage Control via LLM-RL Collaboration: A Hybrid Knowledge-Data-Driven Approach Authors:Xu Yang, Chenhui Lin, Xiang Ma, Dong Liu, Ran Zheng, Haotian Liu, Wenchuan Wu View a PDF of the paper titled Two-Stage Active Distribution Network Voltage Control via LLM-RL Collaboration: A Hybrid Knowledge-Data-Driven Approach, by Xu Yang and 6 other authors View PDF Abstract:The growing integration of distributed photovoltaics (PVs) into active distribution networks (ADNs) has exacerbated operational challenges, making it imperative to coordinate diverse equipment to mitigate voltage violations and enhance power quality. Although existing data-driven approaches have demonstrated effectiveness in the voltage control problem, they often require extensive trial-and-error exploration and struggle to incorporate heterogeneous information, such as day-ahead forecasts and semantic-based grid codes. Considering the operational scenarios and requirements in real-world ADNs, in this paper, we propose a hybrid knowledge-data-driven approach that leverages dynamic collaboration between a large language model (LLM) agent and a reinforcement learning (RL) agent to achieve two-stage voltage control. In the day-ahead stage, the LLM agent receives coarse region-level forecasts and generates scheduling strategies for on-load tap changer (OLTC...