A control optimization model for CVaR risk of distribution systems with PVs/DSs/EVs using Q-learning powered adaptive differential evolution algorithm
•A CVaR-based method is used to determine power control value of PVs, EVs and DSs.•Energy risk due to power loss and voltage offset is controlled by optimizing method.•A method based on second-order cone programming is proposed to control CVaR risk.•Q-learning driven adaptive differential evolution...
Saved in:
Published in | International journal of electrical power & energy systems Vol. 132; p. 107209 |
---|---|
Main Authors | , , , , , , , , |
Format | Journal Article |
Language | English |
Published |
Elsevier Ltd
01.11.2021
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | •A CVaR-based method is used to determine power control value of PVs, EVs and DSs.•Energy risk due to power loss and voltage offset is controlled by optimizing method.•A method based on second-order cone programming is proposed to control CVaR risk.•Q-learning driven adaptive differential evolution method is used for faster solution.
Distributed generation and energy storage brings opportunities and risks of distribution systems, such as greater power loss and abnormal voltage fluctuation. A Q-learning powered optimization model for these risk control of is presented in this paper. Considering the uncertainties of output power of distributed generation systems, charging and discharging power of electric vehicles and energy storage devices, a CVaR-based energy risk control model for distribution system with renewable energy is presented to determine the control value of output power of distributed generation systems. Q-learning powered adaptive differential evolution algorithm is used to solve the proposed optimization problem. Second order cone programming is used to simplify the objective function and constraints of the optimization model, and Q-learning driven adaptive differential evolution algorithm is used to enhance the ability of solving, simplify the calculation, and make the solution faster and more stable. The feasibility and applicability of the proposed model and algorithm are verified by simulating IEEE-118 distribution system. |
---|---|
ISSN: | 0142-0615 1879-3517 |
DOI: | 10.1016/j.ijepes.2021.107209 |