Intelligent Decision-Making System for Power Grid Dispatching Based on Generative AI

Authors

  • Yu Liu State Grid Shanxi Electric Power Company Xinzhou Power Supply Company, Xinzhou 030021, Shanxi, China
  • Weihua He State Grid Shanxi Electric Power Company Xinzhou Power Supply Company, Xinzhou 030021, Shanxi, China
  • Jianfei Zhu State Grid Shanxi Electric Power Company Xinzhou Power Supply Company, Xinzhou 030021, Shanxi, China
  • Ran Yang State Grid Shanxi Electric Power Company Xinzhou Power Supply Company, Xinzhou 030021, Shanxi, China
  • Xinrui Yang State Grid Shanxi Electric Power Company Xinzhou Power Supply Company, Xinzhou 030021, Shanxi, China

Abstract

With the current intelligent decision-making system for the power grid (PG), it is difficult for dispatching mechanisms to automatically update theirdecision logic according to the changing PG status and external environment, and to predict new risks in advance and make adaptive adjustments. In this study, an intelligent decision-making system is constructed for PG dispatching to improve itsintelligence and risk response capabilities. This study collected historical load, grid status and meteorological conditions data, sampled them in 5-minute units, and constructed a data set. The CGAN (Conditional Generative Adversarial Network) was used to take grid status, historical load data, meteorological conditions, etc., as conditional inputs, and generates different risk scenario data through adversarial training. CNN (Convolutional Neural Network) was used to extract local features, which were then input into Bi-LSTM (Bidirectional Long Short-Term Memory Network) for sequence modeling and grid risk identification. The identified risk category and current grid status are used as input, and DQN (Deep Q-Network) uses experience replay and ?-greedy strategy to make scheduling decisions. The results show that the average risk identification accuracy rate in various risk scenarios reached 98.1%, and the average identification precision rate reached 97.8%. Compared with the average response time of 7.4s for traditional systems, the average response time of this system was 1.2s. In various risk scenarios, the average supply and demand balance rates of this system and traditional systems were 0.97 and 0.93 respectively.
Therefore, the intelligent decision-making system proposed in this paper can cope with the changing grid status and external environment, accurately identify risks and respond quickly, showing its broad potential in smart grid applications.

Keywords: Power Grid Dispatching, Intelligent Decision-making System, Generative Artificial Intelligence, Conditional Generative Adversarial Networks, Risk Identification

Cite As

Y. Liu, W. He, J. Zhu, R. Yang, X. Yang, "Intelligent Decision-Making System for Power Grid Dispatching Based on
Generative AI", Engineering Intelligent Systems, vol. 34 no. 1, pp. 5-16, 2026.


Downloads

Published

2026-01-01

Issue

Vol. 34 No. 1 (2026): International Journal of Engineering Intelligent Systems

Section

Article

License

The submission of a paper implies that, if accepted for publication, it will not be published elsewhere in the same form, in any language, without the prior consent of the publisher. Before publication, authors are requested to assign copyright to CRL Publishing Ltd. This allows CRL to sanction photocopying, and to authorize the reprinting of issues or volumes according to demand. Authors' traditional rights will not be jeopardized by assigning Copyright in this way, as they retain the right to reuse the material following publication, and to veto third-party publication.