Research On Combined Denoising Algorithm For Large Structure Vibration Based On GNSS Monitoring

The Global Navigation Satellite System(GNSS)has demonstrated unique advantages in the field of large structure deformation monitoring,with features such as high accuracy positioning and real-time dynamic monitoring.However,GNSS monitoring techniques are susceptible to multi-path errors,resulting in poor measurement accuracy.In addition,the limited sampling rate of the receiver hardware makes it difficult to detect high-frequency structural vibration responses.To address these issues,detailed research is conducted in combined noise reduction algorithms,multi-path error mitigation and multi-sensor data fusion.The main research contents and innovations are as follows:(1)A systematic overview of GNSS relative positioning techniques and the sources of error that affect measurement accuracy.Based on the short baseline RTK-GNSS observation model,it is clarified that multi-path error is one of the main factors leading to the degradation of positioning accuracy.The mechanism of multipath effect generation and its characteristics are analyzed in depth,providing a theoretical basis for further research.(2)For effective extraction of multi-path errors in short baseline RTK-GNSS measurement data,a combined algorithm based on improved complete ensemble empirical mode decomposition with adaptive noise(ICEEMDAN)and adaptive wavelet packet thresholding denoising(AWPTD)is designed.The algorithm processes the signal through three stages: modal function decomposition,effective coefficient sieving and adaptive noise reduction to improve the accuracy of noise reduction while preserving the original characteristics of the signal as much as possible.It was analyzed in simulation experiments and outperformed the single algorithm.(3)Given that GNSS sensors do not easily detect high frequency vibration signals,a multi-rate UKF-based weighted data fusion estimation algorithm is improved using multirate sensor data fusion.The algorithm uses the UKF algorithm to perform local estimation of measurements from sensors with different sampling rates,followed by weighted data fusion estimation of the local estimates for each sensor to obtain the optimal estimate with minimum variance.Simulation experiments show that the algorithm can accurately estimate the true state of the system and that the fusion estimation results are better than the local estimation results of each sensor.(4)Research is carried out into the mitigation of multi-path errors for GNSS observations and the multi-sensor monitoring technique of combined accelerometers,and the effectiveness of the improved algorithms in practical applications is verified through the construction of an experimental platform.Firstly,experiments on multi-path error correction in GNSS monitoring are conducted,and the results show that the multipath error correction model established by the ICEEMDAN-AWPTD algorithm has high accuracy,and the accuracy of E,N and U direction coordinates is improved by 49.2%,65.1% and 56.6%,respectively,after multipath correction.After multi-path error correction,multi-sensor data fusion estimation was performed by combining GNSS and accelerometer measurements.Analysis of the measured data showed that the accuracy of the fused estimated X-direction and Y-direction displacements was improved by 17.2% and 23.8% respectively,while addressing the problem of GNSS sensors not easily detecting high frequency vibration signals.