BP-neural-network-technology-based improved method of spherical harmonic function of ionized layer

Abstract

The invention discloses a BP-neural-network-technology-based improved method of a spherical harmonic function of an ionized layer. The method comprises the following steps: S1, obtaining a puncturingpoint geographic longitude, a puncturing point geographic latitude and an electronic content in the vertical direction of the puncturing point of an observation point of a region; S2, on the basis ofa spherical harmonic function model format of an ionized layer, establishing an ionized layer error compensation model and calculating an electronic content of the puncturing point in the vertical direction and a residual value of the ionized layer error compensation model; S3, carrying out model compensation on a spherical harmonic function by using a BP neural network technology and establishinga fusion model; and S4, with the newly established fusion model, predicting a vertical total ionized layer electronic content residual by spherical harmonic function model fitting, thereby improvingthe model compensation precision. According to the invention, the nonlinear vibration characteristics of the delay information of the ionized layer with time is described accurately; the space-time changing characteristics of the regional ionized layer are simulated well; and compared with the pure spherical harmonic function model, the provided spherical harmonic function model enables the precision to be improved.

Description

technical field [0001] The invention relates to the technical field of global navigation systems, in particular to an ionospheric spherical harmonic function improvement method based on BP neural network technology. Background technique [0002] In the related research and application of GNSS ionospheric modeling, the more commonly used ionospheric correction models are generally divided into two categories: one is the empirical model, including the Klobuchar model, IRI model, Bent model, etc. used in GPS. , because the establishment of empirical models is based on long-term data collection, some models have complex structures and poor applicability, and usually cannot meet the requirements when ionospheric model modeling and GNSS positioning accuracy are high; One is a fitted ionospheric model based on high-precision GNSS dual-frequency observation data. Generally, the commonly used ionospheric models include the following three types: polynomial model (POLY), spherical har...

AI Technical Summary

Problems solved by technology

However, the correction accuracy of these three fitting models is still not very high, and cannot meet the requireme

Images