Combined data-driven and knowledge-driven methodology for system inertia estimation based on the successional difference method.
Saved in:
| Title: | Combined data-driven and knowledge-driven methodology for system inertia estimation based on the successional difference method. |
|---|---|
| Authors: | Shi, Linjun1 (AUTHOR) eec@hhu.edu.cn, Zhao, Yijie1 (AUTHOR) 241306020071@hhu.edu.cn, Wu, Feng1 (AUTHOR), Lin, Keman1 (AUTHOR) |
| Source: | International Journal of Electrical Power & Energy Systems. Sep2025, Vol. 170, pN.PAG-N.PAG. 1p. |
| Subjects: | Inertia (Mechanics), Electric power system stability, Empirical research, Forecasting, Neural computers, Research methodology, Estimation theory |
| Abstract: | • Successional difference method improves frequency change rate calculation via non-adjacent data. • CNN-LSTM-Attention combines spatiotemporal features to correct physical model. • Combined driven method enhances accuracy while enabling real-time prediction. With the development of new power system, the inertia level of power system shows a decreasing trend, which poses a threat to operational safety. Therefore, accurate inertia estimation is crucial to ensure power system stability. In order to overcome the large error problem in the traditional physical model, a frequency rate of change calculation method based on the successional difference method is proposed. Additionally, an emerging CNN-LSTM-Attention neural network is introduced to correct the output of the physical model. Firstly, the post-fault system data are collected and the frequency rate of change caused by disturbance is calculated by the successional difference method, which effectively reduces the accidental error and improves the interpretability of the model at the same time. Then, the improved physical model generates preliminary inertia estimates. However, recognizing the estimation inaccuracies in standalone physical modeling, a combined approach is used. The outputs of the physical model and the collected data are fed into the CNN-LSTM-Attention network. This combined data and knowledge driven methodology significantly improves estimation accuracy. Validation experiments using PSASP simulation and MATLAB analysis on IEEE39 demonstrate the superiority of the proposed method. Comparison results with CNN-LSTM, LSTM and CNN confirm that the CNN-LSTM-Attention neural network achieves optimal performance. This combined approach not only enhances inertia estimation precision but also enables real-time capability, providing a novel and effective solution to the inertia estimation problem. [ABSTRACT FROM AUTHOR] |
| Copyright of International Journal of Electrical Power & Energy Systems is the property of Elsevier B.V. and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) | |
| Database: | Engineering Source |
Be the first to leave a comment!
