Abstract:  Ionospheric delay error is one of the most important error sources in global navigation satellite system (GNSS). At present, the most widely used ionospheric delay correction model is the Klobuchar parameter model, but the correction rate of this model can only reach about 60%, which cannot meet the increasing accuracy requirements. The high-precision ionospheric grid data published by the international GNSS monitoring & assessment system (iGMAS) were used as a reference value to calculate and analyze the error of Klobuchar ionospheric model. The results show that the error in the mid latitude region has obvious periodic characteristics. To further improve the correction accuracy of the Klobuchar ionosphere model in the mid-latitude region, a Klobuchar ionospheric error prediction model was established based on particle swarm optimization back propagation (BP) neural network. Error prediction was made by taking the sample data of October 2019 as an example. The results show that accuracy can be improved to about 90% by using the model to compensate for the ionospheric delay error in the mid latitude region.

Key words: Klobuchar ionospheric model, particle swarm optimization, BP neural network, error analysis, error prediction