Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

Reinforcement Learning for Volt- Var Control: A Novel Two-stage Progressive Training Strategy

Conference · · 2022 IEEE Power & Energy Society General Meeting (PESGM)
 [1];  [2];  [2];  [2];  [2];  [2]
  1. North Carolina State University,Department of Electrical and Computer Engineering,Raleigh,NC; North Carolina State University
  2. North Carolina State University,Department of Electrical and Computer Engineering,Raleigh,NC

This paper develops a reinforcement learning (RL) approach to solve a cooperative, multi-agent Volt-Var Control (VVC) problem for high solar penetration distribution systems. The ingenuity of our RL method lies in a novel two-stage progressive training strategy that can effectively improve training speed and convergence of the machine learning algorithm. In Stage 1 (individual training), while holding all the other agents inactive, we separately train each agent to obtain its own optimal VVC actions in the action space: fconsume, generate, do-nothingg. In Stage 2 (cooperative training), all agents are trained again coordinatively to share VVC responsibility. Rewards and costs in our RL scheme include (i) a system-level reward (for taking an action), (ii) an agent-level reward (for doing-nothing), and (iii) an agent-level action cost function. This new framework allows rewards to be dynamically allocated to each agent based on their contribution while accounting for the trade-off between control effectiveness and action cost. The proposed methodology is tested and validated in a modified IEEE 123-bus system using realistic PV and load profiles. Simulation results confirm that the proposed approach is robust and computationally efficient; and it achieves desirable volt-var control performance under a wide range of operation conditions.

Research Organization:
North Carolina State University
Sponsoring Organization:
USDOE Advanced Research Projects Agency - Energy (ARPA-E)
Contributing Organization:
North Carolina State University
DOE Contract Number:
EE0008770
OSTI ID:
2329462
Journal Information:
2022 IEEE Power & Energy Society General Meeting (PESGM), Journal Name: 2022 IEEE Power & Energy Society General Meeting (PESGM)
Country of Publication:
United States
Language:
English

References (3)

A Data-Driven Multi-Agent Autonomous Voltage Control Framework Using Deep Reinforcement Learning journal November 2020
Deep Reinforcement Learning Based Volt-VAR Optimization in Smart Distribution Systems journal January 2021
Two-Stage Volt/Var Control in Active Distribution Networks With Multi-Agent Deep Reinforcement Learning Method journal July 2021

Similar Records

Attention Enabled Multi-Agent DRL for Decentralized Volt-VAR Control of Active Distribution System Using PV Inverters and SVCs
Journal Article · Wed Feb 03 23:00:00 EST 2021 · IEEE Transactions on Sustainable Energy · OSTI ID:1811839

Volt-VAR Optimization in Distribution Networks Using Twin Delayed Deep Reinforcement Learning
Conference · Sun Apr 24 00:00:00 EDT 2022 · 2022 IEEE Power & Energy Society Innovative Smart Grid Technologies Conference (ISGT) · OSTI ID:1894929

Deep Reinforcement Learning Based Volt-VAR Optimization in Smart Distribution Systems
Journal Article · Fri Jul 17 00:00:00 EDT 2020 · IEEE Transactions on Smart Grid · OSTI ID:1762475