Solar flare dichotomy prediction method based on support vector machine

Abstract

The invention discloses a solar flare dichotomy prediction method based on a support vector machine, and the method comprises the steps: carrying out the data preprocessing of an experimental model, and constructing a data set; determining a Spearman grade correlation coefficient, and obtaining a difference formula between grades; analyzing and testing training features of machine learning, carrying out feature selection on an experimental model sample, and verfiying the relation between the features through a Spearman level correlation coefficient; searching a decision boundary by adopting an SVM, and selecting an SVM kernel function; deducing an SVM calculation principle; performing prediction processing on the data by adopting an SVM, and generating a prediction result; and judging the prediction result. According to the method, the solar flare activity behavior prediction model is constructed by taking the machine learning algorithm as a core, modeling and optimization are performed on the preprocessed data through feature extraction and preprocessing of solar flare observation data, and finally, prediction with relatively high accuracy on solar flare behaviors is realized.

Description

technical field [0001] The invention relates to the technical field of solar flare forecasting, in particular to a support vector machine-based solar flare binary classification forecasting method. Background technique [0002] The sun is not a completely static sphere. The sun has an active cycle and a static cycle. An obvious form of solar activity is solar flares. The huge kinetic energy and thermal energy brought by solar flares can have a great impact on the space environment of the earth. huge impact. [0003] When a solar flare erupts, it will release a lot of radiation and high-energy particles. These companions will have a severe impact on the space environment of the earth. Solar flares are an important part of the space weather forecast model; when the radiation and high-energy particles of the flare appear near the earth, atmospheric molecules The ionosphere in the atmosphere will be damaged by radiation, thus losing the ability to reflect radio waves, seriously...

AI Technical Summary

Problems solved by technology

[0003] When a solar flare erupts, it will release a lot of radiation and high-energy particles. These companions will have a severe impact on the space environment of the earth. Solar flares are an important part of the space weather forecast model; when the radiation and high-energy particles of the flare appear near the earth, atmospheric molecules The ionosphere in the atmosphere will be damaged by radiation, thus losing the ability to reflect radio waves, seriously affecting people's daily communication activities; the flow of high-energy particles emitted by flares will contact the upper atmosphere when it reaches the upper atmosphere. The collision of gas particles will disturb the earth's magnetic field, which will also seriously affect the earth's space environment and affect the research work of researchers; to study and improve the function of solar flare prediction is not only to correctly predict space weather, but also to ensure the research of human space activities. At the same time, it is also an important content to reveal the internal laws of the sun and improve human understanding of the sun; therefore, it is necessary to design a two-category prediction method for solar flares based on support vector machines

Images