Telemetry-Robust Safe Zero-Shot Graph Multi-Agent Reinforcement Learning for Active Voltage Control Across Distribution Feeders
Boyin Jin, Siqi Sun, Yun Zhang, Shang Zheng, Hualong Yu · 2026
Read Telemetry-Robust Safe Zero-Shot Graph Multi-Agent Reinforcement Learning for Active Voltage Control Across Distribution Feeders by Boyin Jin, Siqi Sun, Yun Zhang, Shang Zheng, Hualong Yu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13611244.txt.
Description
Boyin Jin and co-authors present a control system for voltage in power distribution networks in this 2026 paper. Keeping voltage in range depends on sensing, deciding, and acting, and that chain must hold as network shape, solar inverter use, and data quality change. Most learning-based controllers are tied to one feeder's layout.
Their system is built to move to a new feeder with no retraining at all. It uses a shared graph-based actor and twin critics that accept varied network sizes, plus a safety screen based on power flow at use time. On a target feeder it never trained on, voltage violations fell from about 7.5 percent under plain control to about 2.7 percent. Stress tests with missing and faulty data held up well.
Catalog record
| CREATOR | Boyin Jin, Siqi Sun, Yun Zhang, Shang Zheng, Hualong Yu |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 59 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |