Font Size: a A A

A Study Of Time-series Analysis Methods Based-on Wavelet Technology For GPS Deformation Monitoring

Posted on:2005-08-20Degree:MasterType:Thesis
Country:ChinaCandidate:S Z ShiFull Text:PDF
GTID:2120360152955775Subject:Geodesy and Survey Engineering
Abstract/Summary:PDF Full Text Request
On the study involved GPS theory and its monitoring technology, several traditional data processing methods, combining the pragmatic example, A time-series analysis method, based on Wavelet reducing-noises, is put forward and the feasibility of Application for GPS deformation monitoring is demonstrated.The application of GPS for deformation monitoring has wide foreground, but GPS technology, as a new monitoring means, has essential difference in the cause and influence of error from those traditional. If using traditional method in GPS monitoring data processing, we may not reach a desirable result. Given Wavelet technology peculiar function in gross error detection and reducing-noises , combining several classical data-processing methods , synthetically studying Wavelet technology with traditional data processing method, the dissertation recommends a improved time-series analysis method based on Wavelet reducing-noises. This paper begins with the study in basic GPS theory, analyses the error origin in course of GPS survey and brings forward some measures to deal with these impacts from it, expounds concisely the work model, appliance actuality, data collection and observational data processing, details the basic theory about several routine data processing methods, mainly involved regression analysis, spectrum analysis, time-series analysis and Kalman filter. At the same time, on in-depth study of Wavelet theory, combining Wavelet technology, the paper puts forward the improved time-series analysis method, by a test, the result verified its notable advancement on forecast.In the example, the rude data-series of vector Z from GPS observationhas a six-centimetre variational range. After processed with one-dimension Wavelet technology and time-series analysis ordinally, In the result, the value of remnants error in first-step forecast is not more than nine millimetres, which is even twelve millimeters without using Wavelet technology, and its forecast accuracy in second or third step is higher than unimproved time-series analysis method, this tells us the feasibility of the new method.
Keywords/Search Tags:GPS, Deformation monitoring, Regression analysis, Spectrum analysis, Time-series analysis, Higher difference, Kalman filter Wavelet analysis, Improved time-series analysis
PDF Full Text Request
Related items