Open AccessAn AGC dynamic control method based on DQN algorithm
Author(s) -
Su Jianjun,
Chao Ma,
Shan Li,
Xiaoyü Li,
Bing Zhang,
Wenxue Liu
Publication year - 2020
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/729/1/012009
Subject(s) - computer science , convergence (economics) , curse of dimensionality , artificial neural network , control (management) , automatic generation control , electric power system , algorithm , power (physics) , control theory (sociology) , artificial intelligence , physics , quantum mechanics , economics , economic growth
In order to cope with the incompatibility of traditional AGC control under CPS control performance standard, this paper proposes a hierarchical control framework based on DQN algorithm. It effectively solves the problem of dimensionality disaster of Q learning by using neural network instead of state action pair, and speed up the convergence. Through simulation verification, this method can effectively improve the system CPS control performance index and effectively reduce the system power generation cost.