Improved Hybrid Denoising Method for Dynamic Monitoring of a Super-High-Rise Building Based on a GNSS-RTK Technique

Abstract

To enhance the accuracy of dynamic monitoring based on the global navigation satellite system real-time kinematic (GNSS-RTK) technique, an improved hybrid denoising method was proposed in this study. The improved hybrid denoising method is composed of a Type II Chebyshev high-pass filter, improved complete ensemble empirical mode decomposition with adaptive noise (ICEEMDAN), and wavelet soft threshold denoising. Detrended fluctuation analysis (DFA) was applied as a judgment to identify the noise of GNSS-RTK caused by the multipath effects and the receiver’s internal error in the proposed method. A simulation signal containing noise was designed to evaluate the performance of the proposed method. The signal-to-noise ratio (SNR) obtained by the proposed method was 7.1619 dB, which was the largest among the results obtained by the Chebyshev filter, ICEEMDAN, and the proposed method. The root-mean square error (RMSE) obtained by the proposed method was 3.3 mm, which was the smallest among the results obtained by the Chebyshev filter, ICEEMDAN, and the proposed method. The SNR and the RMSE proved the proposed method outperformed the Chebyshev filter and ICEEMDAN. GNSS-RTK stability experiment revealed the noise of GNSS-RTK caused by the multipath effects and the receiver’s internal error. Compared with the signal obtained by wavelet denoiser, the signal obtained by the proposed method was more uniformly distributed around the ideal signal and had a smaller displacement range in the GNSS-RTK stability experiment. It demonstrated the proposed method was more suitable than wavelet soft threshold denoising for GNSS-RTK monitoring denoising. The proposed method was applied in a super-high-rise building with a structure height of 596.2 m. The results demonstrate the proposed improved hybrid denoising method has a significant effect on noise reduction in monitoring the super-high-rise building based on GNSS-RTK. The results also present that the vertical dynamic deformation of the super-high-rise building is greater than the planar dynamic deformation. The first-order natural frequency obtained from processed monitoring data was 0.1926 Hz. The difference is only 1.37% compared with the result obtained by finite-element analysis (FEA). It proves the reliability of the proposed method in monitoring super-high-rise building based on the GNSS-RTK technique.