Welcome to Francis Academic Press

Academic Journal of Computing & Information Science, 2023, 6(13); doi: 10.25236/AJCIS.2023.061325.

Prediction of Sunspot Activity Cycle Based on Long Short-Term Memory (LSTM) Network Models

Author(s)

Yilin Jiang, Yufei Xiong

Corresponding Author:
Yilin Jiang
Affiliation(s)

Southwest University of Political Science and Law, Chongqing, China

Abstract

This paper addresses the prediction of sunspot activity using cycle feature analysis and Long Short-Term Memory Network (LSTM) models, including LSTM-NN. Analysis of sunspot data since 1700 reveals a 100-year major cycle and an 11-year minor cycle, with findings indicating the current 25th solar cycle began in December 2019 and ends in January 2031, and the 26th cycle from January 2031 to January 2041, both resembling the low activity levels of the 24th cycle. Employing a multivariate construction method, the study combines LSTM with univariate and multivariate predictions in single and multi-time steps, with the optimal strategy determined by the minimum root-mean-square error. The multivariable model predicts the peak of the next cycle in November 2035, lasting until May 2037, a duration of 1 year and 6 months.

Keywords

Sunspot Prediction, Cycle Feature Analysis, Long Short-Term Memory Network (LSTM), Multivariate Model

Cite This Paper

Yilin Jiang, Yufei Xiong. Prediction of Sunspot Activity Cycle Based on Long Short-Term Memory (LSTM) Network Models. Academic Journal of Computing & Information Science (2023), Vol. 6, Issue 13: 178-185. https://doi.org/10.25236/AJCIS.2023.061325.

References

[1] Bi Chenglin, Liu Kuang, Xiang Zheng et al. Research on regional LSTM model of flood forecast [J/OL]. Hydropower Energy Science, 2023,(12):63-67[2023-12-04].

[2] Lu H, Yang F .Research on Network Traffic Prediction Based on Long Short-Term Memory Neural Network[C]//2018 IEEE 4th International Conference on Computer and Communications (ICCC).IEEE, 2018.DOI:10.1109/CompComm.2018.8781071.

[3] Zhu Xuefeng, Feng Zao, Ma Jun et al. Defect recognition model of Attention Mechanism U-shaped pipeline based on MSCNN-LSTM [J]. Journal of Vibration and Shock, 2023, 42(22):293-302.

[4] Zhang Wanting. Application of neural network method to solar activity index prediction [D]. National Space Science Center, University of Chinese Academy of Sciences, 2022. (in Chinese)

[5] Xu Hengyu. Long-term variation of sunspot asymmetry between north and south based on complex network [D]. Yunnan University of Finance and Economics, 2022.

[6] Wang X X. Simplified structure design of LSTM neural network and its application in effluent BOD prediction [D]. Beijing University of Technology, 2022.

[7] Yu N. Sunspot time series prediction based on multi-scale decomposition and particle swarm optimization of ELM [D]. Wuhan University of Science and Technology, 2022.

[8] Chen W. Correlation analysis of sunspot activity and sea level temperature [D]. Shanghai Normal University, 2022. (in Chinese)

[9] Ma F, Wang H X. Sunspot prediction based on NMPTS model [J]. Science and Technology Innovation, 2022, (05):189-192.

[10] Lv T , Wang Y , Yuan J ,et al. Prediction and Application of Network Business Traffic based on LSTM[J].IOP Publishing Ltd[2024-01-07]. DOI:10.1088/1742-6596/2289/1/012011.