Research On Data Denoising Method Of GPS Deformation Monitoring Data Based On Wavelet Analysis

In recent years, with the rapid progress of3S technology, deformation monitoring technology is constantly moving forward. Because of the advantage of rapid positioning, high precision, automation, all-weather, without having to be visible between stations and simultaneous determination of the three-dimensional coordinates of stations, GPS technology is widely used in deformation monitoring. GPS singal encounters variety of disturbances coming from error sources, such as GPS receiver’s random noise, ionosphere refraction, troposphere refraction, multi-path effect, etc. Therefore, it is still a research with important theoretical significance and application value for GPS technology to find these errors, reduce them and to get precise deformation observation results in slope monitoring of reservoir area in plateau mountainous area, under the environment affected by obstruction and multi-path etal.Because of its superior ability to de-noising and time-frequency localization analysis, wavelet analysis has been widely applied to the signal processing. Signal can be decomposed into different scales using wavelet analysis. Then we can get the signal component values on each scale, what’s more, the reconstructed signal is also not distorted. The signal of GPS deformation monitoring contains various frequency components, effective information concentrated in low frequency and interference information distributed in high frequency. To improve the accuracy of deformation monitoring, we can eliminate the interference information distributed in high frequency and extract useful information in low frequency using wavelet analysis.In the thesis, the author use wavelet theory as tool, researched how to denoise the GPS slope monitoring data of reservoir area in plateau mountainous area by wavelet analysis method. Using the VC++language and MATLAB as a platform to develop a GPS rinex data pre-process software based on wavelet analysis. In this paper, the author research the optimal method to denoise the GPS slope monitoring data of reservoir area in plateau mountainous area, in order to eliminate the effects of GPS receiver’s random noise and multi-path, and then to improve the quality of baseline solution and the result of network adjustment. Finally, combined with observations of actual engineering to validate the denoising effect of the method proposed in this paper using GPS data processing software. The main research works in this thesis are as follows:(1)This paper summarized the present situation and the research trend of deformation monitoring by GPS technology, what’s more, it has discussed the progress in wavelet theory and the application situation in survey area, and pointed out some questions.(2)This article described the development process from the Fourier transform to wavelet analysis, and introduced the basic theory of wavelet analysis and the maximum decomposition scale.(3)The paper has researched the noise characteristic of GPS signal, using wavelet decompositong method to extract the high-frequency part of raw satellite observations signal, the single difference observed value and the double difference observed value wavelet decomposition. What’s more, the article has conducted the research to the wavelet filter de-nosing method, mainly researched the the de-noising based on wavelet threshold method. Finally this paper has discussed the indicator of wavelet filter effect evaluating.(4)Programmed a GPS rinex data pre-process software based on wavelet analysis, using VC++language and MATLAB as a platform.(5)The author designed four experiments to validate the denoising effect of the GPS rinex data pre-process software based on wavelet analysis, applying the the GPS rinex data pre-process software to pre-process the GPS rinex data of deformation monitoring in Ahai hydropower reservoir area.Chapter V of this paper is the case verification. The experimental results show that to denoise the GPS rinex data with wavelet analysis method has better de-noising effect, using the appropriate wavelet function, decomposition level, the principles of threshold value, the threshold function and threshold reprogramming method can largely eliminate the influence of GPS receiver’s random noise and multi-path effects of the GPS observation data, and greatly improved signal to noise ratio, thereby improving the accuracy of baseline solution and get better net adjustment results.