A new machine learning-based approach for cross-region coupled wind-storage integrated systems identification considering electricity demand response and data integration: A new provincial perspective of China

Xidong Zheng, Sheng Zhou and Tao Jin

Energy, 2023, vol. 283, issue C

Abstract: Faced with the growing renewable energy requirements, there is increased interest in cross-region of large-scale renewable energy market, which provides an alternative path for building sustainable power systems. Critically, the development of a renewable energy-dominated electricity market is an important way to achieve global climate goals and energy conversion. Despite improved achievement of electricity demand response (EDR) market, only limited development in the coordination capacity and development scale of EDR are dominated by renewable energy. Moreover, the normal operation and work efficiency of the system is greatly affected due to data transmission errors and other human factors. Thus, it is of great importance for the State Grid to accurately identify the data sources and realize the interactive development of cross-region power systems. This paper presents a new cross-region (province) electricity demand response (CR-EDR) model in China for large-scale renewable energy participating in EDR market. This model is applied to three provinces in China based on renewable energy, and fully consider how wind power is integrated with EDR in terms of operation, grid connection and optimization. Presently, China is in the initial stage of development, and the identification of data anomalies is inevitable for the CR-EDR. To solve this problem, a novel machine learning (ML)-based approach is proposed for effective identification of EDR. By calculating wind power output and customers’ EDR results, reliable feature sequences can be obtained. Finally, all the feature sequences are uploaded to cross-region system operators (CR-SO) for classification and identification by ML. In the case studies and discussions, we integrated all feature sequences to CR-SO for identification and present a novel process approach to implement EDR, the random forest (RF) enables 100% accuracy for training set and 99.7685% for testing set with small training samples. Compared with RF, support vector machine only achieves 79.1667% for testing set with small training samples. With accurate RF identification results, the stable operation capability and management level of the system can be effectively improved. The proposed methodology creates a new provincial perspective of China and the establishment of CR-EDR, which provides a theoretical and methodological guidance for all countries and regions to develop CR-EDR.

Keywords: Cross-region (province) electricity demand response; Large-scale renewable energy; Coupled wind-storage integrated systems; Machine learning-based approach; Data identification (search for similar items in EconPapers)
Date: 2023
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0360544223025355
Full text for ScienceDirect subscribers only

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eee:energy:v:283:y:2023:i:c:s0360544223025355

DOI: 10.1016/j.energy.2023.129141

Access Statistics for this article

Energy is currently edited by Henrik Lund and Mark J. Kaiser

More articles in Energy from Elsevier
Bibliographic data for series maintained by Catherine Liu ().