Solar activity forecasting method based on spatial feature information of solar observation data

Abstract

The invention discloses a solar activity forecasting method based on spatial feature information of solar observation data, which comprises the following steps of: firstly, respectively training two dichotomy classifiers by using image data of a grade i and a grade j to obtain dichotomy models of different network structures and trained parameters thereof; then loading the obtained dichotomy model and the trained parameters thereof, and obtaining the probability that one picture belongs to the class i by using a classifier trained by the class i data and the class j data; according to the method, an automatic data collection tool is implanted, the data collection period is greatly shortened, the collected data size is greatly increased, the grade of solar flare outbreak is predicted through the convolutional neural network, the prediction effect is good, the prediction success rate is increased through the One-Versus-One scheme, and the prediction efficiency is improved. The multi-classification prediction TSS indexes of the model are 0.7035 + / -0.0703, 0.4893 + / -0.0499, 0.4321 + / -0.2228 and 0.4363 + / -0.3308 respectively, compared with the current international optimal performance index result, the performance is equivalent, and the method has the advantages of being scientific and reasonable, high in applicability, good in effect and the like.

Description

technical field [0001] The invention relates to the technical field of solar activity forecasting, in particular to a solar activity forecasting method based on spatial feature information of solar observation data. Background technique [0002] "Everything grows by the sun", the sun never stops releasing light and heat, providing the proper environment and energy source necessary for human beings and all other life on earth to survive. Because it is inseparable from human production and life, people have a simple desire to deeply understand the nature of the sun. In the process of continuous exploration of the sun, people have discovered many solar activity phenomena, such as flares, coronal mass ejections (CME), sunspots, light spots, etc. These solar activities can cause space weather changes, and severe space weather disturbances will affect space and Ground technical systems (such as satellites, communication networks, power facilities, etc.) cause harm, cause huge eco...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to overcome the deficiencies of the prior art, in order to better and effectively solve the problem that the current prediction accuracy cannot meet the actual needs of space activities for space weather forecasting, and to provide a solar activity forecast based on the spatial feature information of solar observation data method

Images