Research And Application Of Solar Flare Identification And Recording System

Flares are one of the manifestations of solar activity.Although it has not yet caused devastating damage to the earth,it may cause signal interference or even interruption of communication systems on Earth in a short period of time.In order to prevent the impact of flares,many countries in the world have carried out scientific research on flares,but at present,the research on flares at home and abroad is mainly based on forecasting and forecasting,and there is little research on flare identification.The definition of flare on the website is also very simple,and its application is not universal.In view of the current research status,this research will be based on the outlier detection algorithm commonly used in the industrial field,combined with the X-ray flow data provided by the satellite website,to carry out the research on the GOES and RHESSI flare identification algorithms in the time domain,and develop a new algorithm based on the simulated data.A flare automatic identification and recording software.In addition,this research also proposes a method to identify flares in the frequency domain by using wavelet decomposition and utilizing the characteristic fluctuations of the raw data on high-frequency components during flares.Firstly,taking the structure of the sun as the starting point,this thesis summarizes the physical background knowledge of solar activity phenomena,leads to the solar flare,the research object of this thesis,briefly describes the characteristics of the flare and its impact on the earth,and introduces in detail the two X-ray observation satellites GOES and RHESSI.According to the current research status at home and abroad,the purpose and significance of this study are determined.Then,the sources and acquisition methods of GOES and RHESSI satellite data are introduced in detail,and the preprocessing methods are designed according to the data characteristics.After data preprocessing,a recognition algorithm of GOES flares is proposed through statistics,analysis and summary of related parameters of GOES flares.Because RHESSI data is more complex than GOES data,this study proposes a RHESSI flare recognition algorithm based on 3σ criteria and iForest clustering algorithm through a large number of experiments.The two algorithms can complete the flare identification of GOES data and RHESSI data respectively,and save the identified flare list in the specified path.Next,this thesis proposes a method for RHESSI flare identification in the frequency domain.Through research,it is found that after wavelet decomposition of the original data containing flares,characteristic fluctuations will appear in the high-frequency coefficient image corresponding to the flare stage,but there is no such feature in the quiet period,so the characteristic part on the high-frequency coefficient image is corresponding.The flare part,and then the feature part is further enlarged by calculating the short-term energy for easy identification.After the algorithm design is completed,the recognition results of the algorithm are displayed through the renderings.Finally,in order to facilitate the application in different scenarios,this thesis develops two versions of the automatic identification system for the project simulation data.First of all,the basic version(V1.0)is introduced.This version uses PyQt5 to develop a GUI interface.Although the interface is simple,it can realize the core functions of the system.Later,considering the portability of future research work and the professionalism of GUI interface for scientific research,the basic version was improved and optimized with reference to RHESSI software on the IDL platform,and an improved version(V2.0)was developed.The system developed in this research can complete the task of flare identification based on the simulated data provided by the project,which provides a reference for the future ASO-S/HXI to develop a flare identification algorithm based on X-ray flux.