Reactive Power Optimization for Transient Voltage Stability in Energy Internet via Deep Reinforcement Learning Approach

Energies

Junwei Cao¹  Haochen Hua¹  Wanlu Zhang¹  Zeqing Xiao¹ 

[1] Research Institute of Information Technology, Beijing National Research Center for Information Science and Technology, Tsinghua University, Beijing 100084, China;

关键词: energy Internet;    convolutional neural network;    decision optimization;    deep reinforcement learning;   

DOI  :  10.3390/en12081556

来源: DOAJ

【 摘 要 】

The existence of high proportional distributed energy resources in energy Internet (EI) scenarios has a strong impact on the power supply-demand balance of the EI system. Decision-making optimization research that focuses on the transient voltage stability is of great significance for maintaining effective and safe operation of the EI. Within a typical EI scenario, this paper conducts a study of transient voltage stability analysis based on convolutional neural networks. Based on the judgment of transient voltage stability, a reactive power compensation decision optimization algorithm via deep reinforcement learning approach is proposed. In this sense, the following targets are achieved: the efficiency of decision-making is greatly improved, risks are identified in advance, and decisions are made in time. Simulations show the effectiveness of our proposed method.

【 授权许可】

Unknown