Super-Resolution Reconstruction of Power System Data via Weather-Aware Self-Attention TimeGAN
- 6 February 2026
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industry Applications
- Vol. 62 (4) , 5963-5973
Abstract
High-proportional renewable energy integration and rapid installation of new type loads, such as electric vehicles (EVs), into the power system are challenging the monitoring and operation of the power system. Different from the conventional system operation requirements, high-quality and real-time power system data become essential for system operators on key power equipment monitoring, power flow dispatching, and operation optimization. However, due to equipment limitations, bandwidth, storage and other factors, power system data is usually collected at lower sampling frequencies. Therefore, super-resolution reconstruction of these data is necessary to obtain higher frequency data to meet the system operation requirements. In this paper, a super-resolution reconstruction method based on weather-aware self-attention time series generative adversarial network (WA-TGAN) is proposed. The proposed method considers the influence of weather impacts and introduces the self-attention mechanism. The developed generator can more effectively capture long distance dependencies in power system data and improve the quality of generated data, which could help the operators obtain higher resolution and more realistic data and provide high quality data samples for tasks such as new energy forecasting and power quality assessment. The super-resolution reconstruction method proposed in this paper is validated based on actual distribution system data from northern China and publicly available datasets from Austria. The experiment results are validated through indicators such as reconstruction error and timing characteristic evaluation and compared with traditional super-resolution reconstruction methods. The experiment results indicated that the proposed method has better performance compared to the traditional methods.Keywords
Funding Information
- National Natural Science Foundation of China (52207119)
- National Natural Science Foundation of China (U23B20116)
- Young Elite Scientists Sponsorship Program
- CAST (2023QNRC001)
- Shandong Provincial Natural Science Foundation, China (ZR2025QB46)
- Shandong Province Youth Entrepreneurship and Technology Support Program for Higher Education Institutions (2024KJH170)
This publication has 23 references indexed in Scilit:
- Super-Resolution Reconstruction Method for Power System Data Considering Weather Impacts based on Self-Attention TimeGAN2025
- Generative AI for Analog Integrated Circuit Design: Methodologies and ApplicationsIEEE Access, 2025
- Industrial Internet Security Situation Assessment Method Based on Self-Attention Mechanism2024
- Research on Molecular Property Prediction Model Base on Random Forest and LSTM Network2024
- Data Pre-Processing Method for Distribution Networks Based on the CNN-LSTM Considering the Spatial Properties2024
- A Power Data Generation Method Based on CNN-LSTM-GAN for Cable Transmission Line Assessment2023
- Integration of smart grid with renewable energy sources: Opportunities and challenges – A comprehensive reviewSustainable Energy Technologies and Assessments, 2023
- Improved Generative Adversarial Network-Based Super Resolution Reconstruction for Low-Frequency Measurement of Smart GridIEEE Access, 2020
- High Quality Training Set Collection using Generative Adversarial Network2020
- MFZ-KNN — A modified fuzzy based K nearest neighbor algorithm2015