A temporal-domain solar radiation pressure model for BDS-3 MEO orbit prediction

Orbit prediction (OP) for global navigation satellite system (GNSS) satellites is crucial for current location-based Internet of Things (IoT) applications. However, the prediction accuracy is subject to the precision of solar radiation pressure (SRP) model. This paper proposes a novel, time-varying a-priori SRP model for the BeiDou-3 navigation satellite system (BDS-3) medium Earth orbit (MEO) satellite OP, designated A-ECOM. Utilizing the extended CODE (Center for Orbit Determination in Europe) orbit model (ECOM) as the background model, the time series of SRP parameters from 2019 to 2021 are obtained through fitting precise ephemerides; their primary frequencies are then characterized using the fast Fourier transform (FFT). Based on this analysis, A-ECOM is constructed for each satellite, represented by a combination of linear and periodical terms. The OP validation of A-ECOM is conducted with different fitting arc lengths throughout 2022. When predicting BDS-3 orbits with a fitting arc length shorter than 1 d, the utilization of A-ECOM with a relatively tight constraint significantly improves the orbit accuracy compared to the classical ECOM model. For instance, a long-term OP (14-d) with a 4-h fitting arc length reveals a dramatic reduction in the average signal-in-space ranging error (SISRE) from 250.36 m to 7.17 m for MEO satellites manufactured by China Academy of Space Technology (MEO-CAST) and from 273.84 m to 9.36 m for MEO satellites manufactured by Shanghai Engineering Center for Microsatellites (MEO-SECM). Furthermore, this model features an advantage for long-term OP as it captures the temporal variations in SRP parameters, thereby generating dynamic SRP parameters during orbit fitting and prediction. For the 14th-d OP with a long fitting arc length of 5 d, the average SISRE of MEO-CAST exhibits a significant reduction from 3.69 m to 1.90 m, while that of MEO-SECM is reduced from 2.51 m to 2.06 m, compared to the ECOM model.