Solar Flare Prediction Based on the Fusion of Multiple Deep-learning Models
Solar Flare Prediction Based on the Fusion of Multiple Deep-learning Models
复制标题
基于多种深度学习模型融合的太阳耀斑预测
DOI：
10.3847/1538-4365/ac249e
复制
发表时间：
2021-12
期刊：
影响因子：
--
通讯作者：
Zhiping Wu
中科院分区：
文献类型：
--
作者：
Rongxin Tang;Wenti Liao;Zhou Chen;Xunwen Zeng;Jing-song Wang;Bingxian Luo;Yanhong Chen;Yanmei Cui;Meng Zhou;Xiaohua Deng;Haimeng Li1;Kai Yuan;Sheng Hong;Zhiping Wu
太阳耀斑的形成机制及其预测一直是太阳物理学研究的难点。传统的预测方法是人工建立太阳活动区的观测值与太阳耀斑之间的统计关系，不能充分利用观测数据中包含的与太阳耀斑有关的信息。本文首先利用太阳黑子群的实测磁图和磁特征参数驱动神经网络学习预测模型，预测太阳耀斑。该预测融合模型基于深度神经网络、卷积神经网络和双向长短期记忆神经网络，可以预测未来24或48 h内太阳黑子群是否会发生M级或C级以上的耀斑事件，并使用真实的技能统计量（TSS）和F1得分来评估融合模型的性能。实验结果表明，该融合模型能够充分利用太阳耀斑的相关信息，联合收割机能够综合各独立模型的优点，捕捉太阳耀斑的演化特征，比传统的统计预测模型或单一的机器学习方法有更好的性能。我们还提出了两个框架，即F1_FFM和TSS_FFM，分别优化F1评分和TSS评分。交叉验证结果表明，它们在F1得分和TSS得分上各有优势。
Solar flare formation mechanisms and their corresponding predictions have commonly been difficult topics in solar physics for decades. The traditional forecasting method manually constructs a statistical relationship between the measured values of solar active regions and solar flares that cannot fully utilize the information related to solar flares contained in observational data. In this article, we first used neural-network methods driven by the measured magnetogram and magnetic characteristic parameters of the sunspot group to learn the prediction model and predict solar flares. The prediction fusion model is based on a deep neural network, convolutional neural network, and bidirectional long short-term memory neural network and can predict whether a sunspot group will have a flare event above class M or class C in the next 24 or 48 hr. The real skill statistics (TSS) and F1 scores were used to evaluate the performances of our fusion model. The test results clearly show that this fusion model can make full use of the information related to solar flares and combine the advantages of each independent model to capture the evolution characteristics of solar flares, which is a much better performance than traditional statistical prediction models or any single machine-learning method. We also proposed two frameworks, namely F1_FFM and TSS_FFM, which optimize the F1 score and TSS score, respectively. The cross validation results show that they have their respective advantages in the F1 score and TSS score.
登录
查看更多内容
影响因子：
1.8
作者：
Xin Huang;Huaning Wang
通讯作者：
Xin Huang;Huaning Wang
DOI：
10.1086/595550
发表时间：
2008-11
期刊：
The Astrophysical Journal Letters
影响因子：
--
作者：
G. Barnes;K. Leka
通讯作者：
G. Barnes;K. Leka
DOI：
10.1086/421261
发表时间：
2004-03
期刊：
The Astrophysical Journal
影响因子：
--
作者：
M. Wheatland
通讯作者：
M. Wheatland
影响因子：
1.8
作者：
Jinfu Liu;Fei Li;J. Wan;Daren Yu
通讯作者：
Jinfu Liu;Fei Li;J. Wan;Daren Yu
影响因子：
3.3
作者：
A. McCloskey;Peter T. Gallagher;D. S. Bloomfield
通讯作者：
A. McCloskey;Peter T. Gallagher;D. S. Bloomfield