Research of Satellite Clock Error Prediction Based on RBF Neural Network and ARMA Model

As the main error sources of the observation data, the precision of prediction model has a direct effect on the performance of navigation system. Considering that the clock error was composed of trend part and random component, an integrated model was proposed, which was based on RBF neural network and ARMA. The trend was modeled using the RBF neural network, while the random part by the ARMA model, and last added them to the predicted results. The simulation results validate the feasibility and the better performance of the integrated method through an example by using the precise IGS clock data.