关键词: 钟差建模与预报, 星间链路, 时间同步体制, 组合模型, Lomb-Scargle谱分析

Abstract: BeiDou Navigation Satellite System (BDS) achieves autonomous orbit determination and time synchronization capabilities through inter-satellite link (ISL) technology,enabling the determination and predictive modeling of satellite clock biases using ISL measurements during autonomous operation.Addressing discontinuous and non-uniform data characteristics in BDS ISL measurements,this study proposes a hybrid modeling approach integrating a quadratic polynomial (QP) model with the Lomb-Scargle (LS) periodic correction algorithm to effectively decompose clock biases into trend,periodic,and stochastic components.Validations using BeiDou-3 ISL measurements demonstrate that the QP+LS fusion model significantly enhances prediction accuracy:3-hour predictions achieve an RMS better than 0.128 ns,and 24-hour predictions an RMS better than 0.442 ns,representing average improvements of 29.68% and 17.60% respectively over the conventional QP+FFT algorithm.These results provide critical technical references for clock bias prediction in BDS autonomous operations and clock modeling under constrained measurement data conditions.

Key words: clock bias modeling and prediction, inter-satellite link, time synchronization mechanism, hybrid model, Lomb-Scargle spectral analysis