Short-term satellite clock bias forecast based on complementary ensemble empirical mode decomposition and quadratic polynomial. Issue 389 (4th March 2023)

• Record Type: Journal Article
• Authors: Tan, Xiaorong; Xu, Jiangning; He, Hongyang; Chen, Ding; Liang, Yifeng; Wu, Miao
• Is Part Of: Survey review. Volume 55:Issue 389(2023)
• Page Start: 127
• Page End: 136
• Publication Date: 2023-03-04
• Subjects: Satellite clock bias forecast -- Empirical mode decomposition -- Quadratic polynomial -- Permutation entropy -- Pearson correlation coefficient; Surveying -- Periodicals; Great Britain -- Surveys -- Periodicals; 526.9
• DOI: 10.1080/00396265.2022.2025714
• Languages: English
• ISSNs: 0039-6265

Abstract

Abstract : The nonlinear and nonstationary characteristics of satellite clock bias (SCB) have a harmful effect on the accuracy and stability of SCB forecast. To eliminate the influence of nonlinearity and non-stationarity, a hybrid forecast model was constructed that combines complementary ensemble empirical mode decomposition (CEEMD) and quadratic polynomial (QP), called CEEMD-QP. First, the SCB sequence is decomposed into several intrinsic mode function (IMF) components and one residual term by CEEMD. Second, permutation entropy (PE) and the correlation coefficient are used to quantitatively determine the IMF component with more noise and weak correlation with the original SCB signal. Finally, the SCB is reconstructed with the IMFs and residual components, and QP model is used to fit and forecast the clock bias. We adapt the observation part of ultra-rapid precise SCB data of GPS provided by IGS to forecast experiments. The results show that the CEEMD-QP method has obvious advantages of forecast accuracy and stability in short-term forecast.