Efficient and Stable Learning for Distribution Network Operation: A Model-Based Reinforcement Learning Approach

Dong Yan; Zhan Shi; Xinying Wang; Yiying Gao; Tianjiao Pu; Jiye Wang

doi:10.17775/CSEEJPES.2023.09100

欢迎来到能源电力期刊网！

English Version 期刊登录

您当前的位置：

首页 >

文章列表页 >

Efficient and Stable Learning for Distribution Network Operation: A Model-Based Reinforcement Learning Approach

Regular Papers | 更新时间：2026-10-15

- Efficient and Stable Learning for Distribution Network Operation: A Model-Based Reinforcement Learning Approach
- Efficient and Stable Learning for Distribution Network Operation: A Model-Based Reinforcement Learning Approach
- 中国电机工程学会电力与能源系统学报(英文) 2025年11卷第3期页码：1080-1092
- 作者机构：
  
  1. China Electric Power Research Institute,Beijing,China
  2. State Grid Digital Technology Holding Co., Ltd.,Beijing,China
- 作者简介：
- 基金信息：
- DOI：10.17775/CSEEJPES.2023.09100
  中图分类号：
- 纸质出版：2025-05-01
- 稿件说明：
移动端阅览
Dong Yan, Zhan Shi, Xinying Wang, 等. Efficient and Stable Learning for Distribution Network Operation: A Model-Based Reinforcement Learning Approach[J]. 中国电机工程学会电力与能源系统学报(英文), 2025,11(3):1080-1092.

Dong Yan, Zhan Shi, Xinying Wang, et al. Efficient and Stable Learning for Distribution Network Operation: A Model-Based Reinforcement Learning Approach[J]. CSEE Journal of Power and Energy Systems, 2025, 11(3): 1080-1092.
Dong Yan, Zhan Shi, Xinying Wang, 等. Efficient and Stable Learning for Distribution Network Operation: A Model-Based Reinforcement Learning Approach[J]. 中国电机工程学会电力与能源系统学报(英文), 2025,11(3):1080-1092. DOI： 10.17775/CSEEJPES.2023.09100.

Dong Yan, Zhan Shi, Xinying Wang, et al. Efficient and Stable Learning for Distribution Network Operation: A Model-Based Reinforcement Learning Approach[J]. CSEE Journal of Power and Energy Systems, 2025, 11(3): 1080-1092. DOI： 10.17775/CSEEJPES.2023.09100.

摘要

Abstract

This paper discusses the application of deep reinforcement learning (DRL) to the economic operation of power distribution networks

a complex system involving numerous flexible resources. Despite the improved control flexibility

traditional prediction-plus-optimization models struggle to adapt to rapidly shifting demands. Modern artificial intelligence (AI) methods

particularly DRL methods

promise faster decision-making but face challenges

including inefficient training and real-world application. This study introduces a reward evaluation system to assess the effectiveness of various strategies and proposes an enhanced algorithm based on the Model-based DRL approach. Incorporating a state transition model

the proposed algorithm augments data and enhances dynamic deduction

improving training efficiency. The effectiveness is demonstrated in various operational scenarios

showing notable enhancements in rationality and transfer generalization.

关键词

Keywords

references

27. OASIS. ( 2021 ). California ISO open access same-time information system . [Online ] . Available: http://oasis.caiso.com/mrioasis/logon.do.

浏览量

219

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

Physics-Informed Generative AI for Power Outage Scenario Synthesis

Strategic PV Electricity Bidding in Spot Markets Under Forecast Uncertainty

Multi-Agent Deep Reinforcement Learning Approach for Temporally Coordinated Demand Response in Microgrids

A super-resolution perception-based incremental learning approach for power system voltage stability assessment with incomplete PMU measurements

Automatic Programming for the Optimal Power Flow in Distribution Networks: An Edge-side Adaptive Computing Approach Based on Large Language Models