基于改进灰色ARMA模型的卫星钟差短期预报研究

RESEARCH ON SHORT-TERM PREDICTION OF SATELLITE CLOCK ERRORS BASED ON IMPROVED GREY ARMA MODEL

导航卫星钟差的精度直接影响导航定位性能。针对卫星钟差由趋势项和随机项组成的特点,提出一种改进灰色模型和ARMA模型的钟差预报组合模型。对传统灰色模型进行改进并建立趋势项预报模型,提取钟差随机项建立ARMA模型,最后将预报结果相加。在算例中采用IGS提供的精密钟差进行预报,仿真结果表明钟差精度较高。

As one of the key factors for realizing real time precise point positioning, the precision of clock er- rors prediction has direct effect on the capability of navigation system. Through considering that the clock errors are composed of trend item and random item, a integrated model is proposed, which combines improved Grey with AR- MA model. Using the improved Grey model to establish trend item model at first, and then we extract the random i- tem to build the ARMA model, add them to the predicted results at last. The simulation results verify the feasibility and validity of this integrated method through an example using the precise clock errors from IGS （ International GPS Service）.