Abstract：In order to improve the effectiveness and reliability of noise reduction from GNSS deformation monitoring data, we propose the complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) method. First, the GNSS deformation sequence is decomposed into several characteristic modal functions by CEEMDAN. Second, we introduce the permutation entropy theory to determine the high and low frequency boundary value K, then use wavelet analysis to denoise the high frequency component. We reconstruct the denoising sequence with the low frequency component after denoising. Finally, through simulation data and measured slope GNSS deformation monitoring data, we compare and analyze CEEMDAN, EMD and wavelet denoising methods using signal-to-noise ratio, root mean square error, correlation coefficient and other indicators. The results show that CEEMDAN is superior to EMD and wavelet denoising methods, proving the effectiveness and reliability of the method proposed in this paper.
