Xing Su / Shandong University of Science and Technology
The ionosphere, as an important part of the space environment, can reflect, refract, and scatter radio signals, which can change the propagation speed and direction of navigation signals, thus causing ionospheric delay, which is one of the important error sources in GNSS measurements. The BDS Klobuchar (BDSKlob) and BDGIM model parameters are broadcast by the BDS in the broadcast ephemeris to correct ionospheric delay errors, which can meet the basic navigation and positioning needs of users. However, in the face of the growing demand for autonomous positioning and navigation accuracy, it is necessary to further improve the accuracy of the model and reduce the impact of the space environment on the positioning. In this work, a Back Propagation (BP) neural network optimized by Artificial Bee Colony Algorithm (ABC) is used to compensate for the error prediction of the BDS broadcast ionosphere model from 7 to 13 September 2021. For BDSKlob and BDGIM, a number of grid points in the Chinese region and worldwide are selected for experimental analysis. BDSKlob and BDGIM respectively selected several grid points in China and the world for experimental analysis. The results show that the prediction compensation of the BDS broadcast ionosphere model errors using the ABC-BP neural network can achieve better accuracy results. For BDSKlob, the model correction rate improved to 81.66% in China after using the predicted values to compensate for the model values. For BDGIM, the accuracy was significantly improved in the global mid and high latitudes, with model correction rates of 74.25%, 82.05%, and 82.13% for the high, mid, and low latitudes respectively after compensation. The experimental results show that it is reasonable and feasible to use ABC-BP neural network to model the error prediction of the BDS broadcast ionosphere model, and the model accuracy can be improved after error compensation.