Skip to content

Fast-Converged Deep Reinforcement Learning for Optimal Dispatch of Large-Scale Power Systems under Transient Security Constraints

Notifications You must be signed in to change notification settings

xxh0523/FCDRL_TSCOPF

Repository files navigation

FCDRL_TSCOPF

This repository contains the whole implementation of the Fast-Converged Deep Reinforcement Learning for Optimal Dispatch of Large-Scale Power Systems under Transient Security Constraints. The code is developed using Py_PSOPS.

Preparation

Just download the whole repository. The code can run on Windows and Linux platforms.

Requirements

Install the following packages before running the code.

conda install qtwebkit
conda install numpy
pip install torch
pip install scikit-opt
pip install ray
pip install timebudget

Agent training

With the default settings of hyperparameters, an agent can be trained with the following command.

python sopf_base.py --training

References

[1] T. Xiao, Y. Chen*, H. Diao, S. Huang, C. Shen, “On Fast-Converged Deep Reinforcement Learning for Optimal Dispatch of Large-Scale Power Systems under Transient Security Constraints,” arxiv

[2] T. Xiao, Y. Chen*, J. Wang, S. Huang, W. Tong, and T. He, “Exploration of AI-Oriented Power System Transient Stability Simulations,” Journal of Modern Power Systems and Clean Energy, vol. 11, no. 2, pp. 401–411, Mar. 2023, doi: 10.35833/MPCE.2022.000099

About

Fast-Converged Deep Reinforcement Learning for Optimal Dispatch of Large-Scale Power Systems under Transient Security Constraints

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published