Application of back-propagation neural networks improved by particle swarm optimization algorithm in ionospheric modeling

View at publisher Summarise with VIRAT
Authors: Cao Lu, Yuan Chen, Linchao Huang, Yang Fang, Liya Ji, Wenpu Wang
Cited by: 0
References: 9

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

This research proposes an improved back-propagation (BP) neural network ionospheric modeling method based on the particle swarm optimization (PSO) algorithm. The primary objective is to enhance the modeling accuracy of total electron content (TEC) in the ionosphere while overcoming the inherent limitations of traditional BP neural networks, specifically their sensitivity to initial weights and slow convergence speed. The proposed method leverages the global search capability of the PSO algorithm to optimize the initial weights and thresholds of the BP neural network, thereby improving the model's convergence performance and prediction accuracy. The model was trained and tested using Global Ionospheric Grid Data (GIM) from the European Centre for Orbit Determination (CODE) spanning the period from 2020 to 2024. Evaluation metrics included the mean residual (AVE), root mean square error (RMSE), and mean relative accuracy (RA). Experimental results indicate that the PSO-BP model effectively establishes ionospheric models across low, medium, and high latitude regions. Specifically, the RA exceeds 95% in low-latitude regions, 91% in mid-latitude regions, and 86% in high-latitude regions. Furthermore, compared to the traditional BP model, the PSO-BP model reduces the RMSE by an average of 0.49 TECU across all test grid points, corresponding to an accuracy improvement of approximately 18.9%. Notably, the performance enhancement is most pronounced in high-latitude regions, suggesting improved robustness in complex environments.

Fields of Study & Subjects

Ionosphere and magnetosphere dynamics Neural Networks and Applications Hydrological Forecasting Using AI

Publication details

Journal / source
International Conference on Algorithms, High Performance Computing and Artificial Intelligence (AHPCAI)
Pages
136–136
Publication date
19 March 2026
Language
EN