实时钟差估计算法实现及精度分析

Realization and Accuracy Evaluation of Real-time Clock Bias Estimation Algorithm

研究并实现了基于非差相位和伪距的实时钟差估计方法，对未知参数采用不同噪声方差的随机游走模型进行约束并采用Kalman滤波进行估计。采用64个IGS监测站的GPS观测数据进行模拟实时钟差解算实验验证算法的正确性和有效性，结果表明，实时钟差的首次收敛时间约需3～4h。收敛之后的卫星钟差相对于IGS最终产品的RMS和STD分别为约0．1ns和优于0．04ns，这一结果与IGS分析中心最终产品的精度相当。

A real-time clock bias estimation algorithm based on undifferenced phase and pseudo range is presented and real- ized. Unknown parameters are limited with random walk model using different noise variances and estimated with the Kalman fil- tering. GPS data from 64 IGS stations were collected to carry out experiment of simulated real-time clock bias estimation to evalu- ate the correctness and effectiveness of algorithm. The results indicate that it took about 3 ~ 4 hours for the first time convergence of real time clock bias. Compared with IGS final clock products, the daily RMS and STD of clock bias after convergence are about 0.1 ns and better than 0.04 ns respectively, equal to that of the final clock products provided by IGS analysis centers.