Application Of Method Of CEEMD Combined With Improved Wavelet Threshold De-Nosing And PCA In GPS Multipath

In GPS measurement of short baseline,carrier phase dual-frequency differential techniques can be used to eliminate ionosphere and troposphere errors,receiver and satellite clock errors,but random noise errors and GPS multipath effect errors cannot be weakened by this method.This article carries out the correlative analysis of the removal of high-frequency random noise and the correction of GPS multipath effect errors.The analysis content is as follows:(1)This article summarizes the research background of GPS multipath effect and the research status at home and abroad and introduce the data post-processing methods of multiple GPS multipath effects in detail.Firstly,the wavelet filtering method is introduced.An improved wavelet threshold de-noising method is proposed for the shortcomings of wavelet hard and soft threshold functions.Experiments show that the improved method has achieved good results.According to the phenomenon of modal aliasing in Empirical Mode Decomposition(EMD),the Complementary Ensemble Empirical Mode Decomposition(CEEMD)is introduced to perform EMD,overall empirical mode decomposition,and CEEMD filter de-noising on simulated signals,the experimental results show that CEEMD filtering method has better de-noising effect.(2)In view of the phenomenon that the low frequency effective signal region still exists noise in the process of CEEMD de-noising,using the improved wavelet threshold function method to de-noise the part of the noise by threshold,the CEEMD combined improved wavelet threshold function de-noising method is proposed.Compared with the traditional CEEMD filtering method and the wavelet soft threshold de-noising method,the experimental results show that the combined de-noising ability is stronger.Then the method is applied to remove high-frequency random noise to better extract and weaken GPS multipath effect errors.The analysis of measured data shows that these methods all play a role in weakening GPS multipath,and CEEMD combined with improved wavelet threshold function has the best de-noising effect.(3)According to repeatability of GPS multipath error was weakened day by day,PCA can be better extracted for GPS multipath effect error,based on several CEEMD denoising methods to remove noise,using PCA method to extract the common error model in the GPS multipath effect error,the GPS multipath effect error is corrected according to the model.Analysis of the results of several other methods by comparison,It shows that the method proposed in this paper can better weaken the GPS multipath effect error.Through experiments,the correlation coefficients between corrected sequence and original deformation sequence are all improved,The combination method of CEEMD combined with improved wavelet threshold and PCA has a greater degree of improvement,It shows that the method has some effect on the correction of GPS multipath effect error in dynamic deformation monitoring.