Constrained Reinforcement Learning-Based Framework for DC Microgrid: Current Sharing and Voltage Regulation
- 1Chongqing Panlong Pumped Storage Power Station Co., Ltd., Chongqing 401452, China
- 2College of Electronic and Information Engineering, Southwest University, Chongqing 400715, China
Abstract
In this paper, a constrained reinforcement learning-based framework is developed for islanded DC microgrids to deal with the issues of voltage regulation and current sharing. First, a new local Q function including the current error and voltage error is designed on the basis of a hyperbolic function. The solution of the HJB equation associated with the local Q function yields an optimal policy. Then, a novel saturation-input policy iterative (SIPI) method is designed to approximate this policy. Its convergence is also analyzed in detail. To implement this SIPI method, an actor–critic control structure is developed, in which the actor network employs an asymmetric bounded function as its activation function to handle asymmetric input saturation. Finally, the effectiveness of the SIPI method is verified through numerical simulations.
Related
- Reinforcement learning for constrained stochastic linear--quadratic control in continuous time
- Distribution-Agnostic Robust Trajectory Optimization via Chance-Constrained Reinforcement Learning
- Rare Event Constrained Reinforcement Learning with Large Deviations
- Chance-Constrained Inspection Viewpoint Sequencing for On-Orbit Servicing via Safe Reinforcement Learning
- A Power Regulation Strategy for Grid-Forming Wind Power Clusters Based on Deep Reinforcement Learning
BibTeX
@article{li2026constrained,
title = {{Constrained Reinforcement Learning-Based Framework for DC Microgrid: Current Sharing and Voltage Regulation}},
author = {Jiawen Li and Hongyu Sun and Dajie Hui and Jianfeng Zhou and Ziwen Shen and Tao Dong},
journal = {Mathematics},
year = {2026},
month = sep,
doi = {10.3390/math14193556},
url = {https://www.mdpi.com/2227-7390/14/19/3556},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Constrained Reinforcement Learning-Based Framework for DC Microgrid: Current Sharing and Voltage Regulation
AU - Jiawen Li
AU - Hongyu Sun
AU - Dajie Hui
AU - Jianfeng Zhou
AU - Ziwen Shen
AU - Tao Dong
PY - 2026
DA - 2026/09/30
JO - Mathematics
DO - 10.3390/math14193556
UR - https://www.mdpi.com/2227-7390/14/19/3556
AB - In this paper, a constrained reinforcement learning-based framework is developed for islanded DC microgrids to deal with the issues of voltage regulation and current sharing. First, a new local Q function including the current error and voltage error is designed on the basis of a hyperbolic function. The solution of the HJB equation associated with the local Q function yields an optimal policy. Then, a novel saturation-input policy iterative (SIPI) method is designed to approximate this policy. Its convergence is also analyzed in detail. To implement this SIPI method, an actor–critic control structure is developed, in which the actor network employs an asymmetric bounded function as its activation function to handle asymmetric input saturation. Finally, the effectiveness of the SIPI method is verified through numerical simulations.
SN - 2227-7390
ER -