•  
  •  
 

Keywords

large‑scale virtual energy storage; new energy; power; prediction error; confidence; scheduling; particle swarm optimization algorithm

Abstract

Large scale virtual energy storage is a large‑scale energy storage system composed of multiple discrete energy storage devices through virtualization technology in the power grid, in order to achieve power balance regulation of the power grid. Because of the randomness, fluctuation and intermittence features of new energy power generation, it is difficult to control the prediction error of new energy power. In order to improve the local consumption level of new energy and reduce the prediction error of new energy power, an optimal scheduling method of large‑scale virtual energy storage to suppress the prediction error of new energy power is proposed. By setting the time resolution of new energy power prediction, the new energy power stabilizing distribution characteristics of large‑scale virtual energy storage are counted, the distribution characteristics of new energy power prediction error are determined, the confidence interval of new energy power prediction error is estimated, the new energy prediction power is included in the power generation plan according to certain confidence degree, the constraint conditions of large‑scale virtual energy storage to stabilize new energy power prediction error are designed, the optimal scheduling model of new energy power prediction error is constructed, and the optimal solution of the model is solved by using particle swarm optimization algorithm. The experimental results show that the proposed method is more sensitive to large‑scale virtual energy storage to stabilize the prediction error of new energy power, with less change in high‑energy load regulation and lower cost, and has remarkable economy and effectiveness.

DOI

10.19781/j.issn.1673-9140.2023.06.018

First Page

167

Last Page

174

Download

Share

COinS