A method of ionosphere forecasting based on vmd and elman neural network

Abstract

The invention discloses an ionospheric forecasting method based on VMD and Elman neural network, which utilizes the combination of variational mode decomposition method and Elman neural network to carry out ionospheric TEC forecasting modeling. The invention utilizes the combination of variational mode decomposition (VMD) and Elman neural network to carry out ionospheric TEC prediction modeling. Considering the nonlinear and non-stationary characteristics of the ionospheric TEC sequence, the VMD algorithm can effectively reduce the complexity of the original sequence and avoid the phenomenon of mode aliasing when preprocessing the original TEC sequence. Compared with the commonly used EMD algorithm, VMD can effectively avoid the occurrence of modal aliasing in the process of data preprocessing, and the performance of filtering and noise reduction is better. The present invention introduces the VMD method to decompose the ionospheric TEC sequence to obtain the corresponding intrinsic mode component (IMF), which provides higher-quality input values ​​for subsequent prediction models.

Description

technical field [0001] The invention relates to ionospheric technology, in particular to an ionospheric forecasting method based on VMD and Elman neural network. Background technique [0002] At present, the ionosphere, as an important part of the near-Earth space environment, has an important impact on the accuracy and real-time performance of navigation and positioning systems and radio communications. As an important parameter of the ionosphere, the total electric content (TEC) of the ionosphere can be accurately predicted to effectively avoid the impact of the ionosphere on people's lives. In order to seek a high-precision ionospheric TEC prediction model, scholars at home and abroad have carried out extensive research and achieved certain results. The ionosphere is disordered, random, and nonlinear in time and space. If the inherent model is selected to predict the electron content of the ionosphere, the accuracy of the obtained TEC can no longer meet the application r...

AI Technical Summary

Problems solved by technology

In the process of data training, this method is very easy to make the model fall into the infinite loop of local minimum, thus avoiding the real optimal value

On the other hand, the structure of the BP neural network is input layer-hidden layer-output layer. Although the hidden layer is connected with the input and output layers during the training process, the single neuron contained in each hidden layer But they are independent of each other, and there is no feedback of information between neurons

The lack of information feedback within this layer further reduces the data processing capability of the network structure

Images