Ionospheric TEC Prediction Based On Wavelet Decomposition And LSTM-Elman Model

The total electron content(TEC)of the ionosphere is an important parameter to describe the number of free electrons in the ionosphere and the characteristics of the ionosphere.Predicting and analyzing the TEC has always been an important topic in the study of the ionosphere.Its changes are closely related to human production and life,and also provide a theoretical and data basis for global navigation and positioning system,aerospace,earthquake and communication.The main research contents are as follows:Firstly,the content and format of TEC data provided by IGS(International GNSS Service)center are introduced,and the influence of geomagnetic and solar activities on the characteristics of ionospheric TEC is analyzed through 2014 and 2018 TEC data.The results show that the ionosphere will change with the change of solar activity,and the magnetic storm will scatter the ionosphere,and the average correlation coefficients between solar activity,geomagnetic activity and ionospheric TEC are 0.25 and 0.18.Then,the principles of wavelet decomposition,long and short term memory(LSTM)network model and Elman neural network model are introduced respectively to determine the optimal level of wavelet decomposition.A neural network combination model based on LSTM-Elman after wavelet decomposition is proposed.Finally,the ionospheric TEC data provided by IGS Center in 2014 and 2018 with the same longitude and latitude but different solar and electromagnetic intensities,the same longitude and latitude but different time periods,the same longitude and time periods but different latitudes,the same latitude and time periods but different longitudes are used for analysis.The original TEC series of the previous 20 days are preprocessed through wavelet decomposition to decompose high-frequency components and low-frequency components,Secondly,LSTM and Elamn are used to predict the high-frequency and low-frequency components respectively,and finally the ionospheric TEC prediction sequence is reconstructed,and the data in the next 5 days are analyzed and studied.The results showed that the RMSE,MAE and RA of the combined model were 0.75 TECu,0.59 TECu and 95.6%,respectively.It is better than two single models,indicating that the combined model has good stability in ionospheric TEC prediction.The paper are 29 figures,22 tables,75 references.