Study On Data Processing Model Of Dynamic Deformation Monitoring By GPS

With more and more high-rise buildings construction, the structure health isconcerned commonly by the people. If the buildings deformation information andextent of the damage can be achieved timely under external factors like typhoon,earthquake and so on, it can make sure the safe operation of the high-rise building.Compared with traditional monitoring tools, GPS featured its high frequency,real-time and non-interference of external environment, which makes GPStechnology to be an important new approach.This paper focuses on the use of GPS in dynamic deformation monitoring, themain contents are as follows.(1) For large pseudorange noise under complex environment, a new type ofcombination algorithm for cycle slip detection and repairis proposed, with theexperiment to demonstrate its effectiveness and practicality. A program to solveshort baseline is designed with the open source code of RTKLib relative positioningmodule, which compared with TRACK through an experiment to verify the highaccuracy.(2) The basic method of troposphere delay is introduced and the change oftemperature, relative humidity, atmospheric pressure and elevation on the commoneffect of tropospheric delay is simulated. Stochastic process method is adopted toestimate the parameters of zenith tropospheric delay added in kalman filter. Twogroups of baseline solving experiment, one with no height difference, the other withheight difference, is designed to analysis the correction effect of ZTD.(3) Correlation for two consecutive days of one station multipath effects isanalyzed from the point view of pseudorange multipath and station coordinatessequence to demonstrate the day repeated model of multipath. Adaptive filter isdesigned to separate two consecutive days coordinates, the result of which relatedpart is multipath error. Through the simulation experiment, the effectiveness ofadaptive filter separating signals can be validated, then testing its effects withmeasured data in multipath error processing. For a single signal or poor correlativedata, while the adaptive filter is no longer applicable, using wavelet analysis toextract multipath effects is proposed. By shaking table experiment, the measurementnoise, vibration signals, multipath cycle effects and the combined effects oflow-frequency multipath with temperature and so on. (4) Taking the data measured in a Hongkong high-rise building during thetyphoon as an example to analyze suitability and effectiveness of the above theory inhigh-rise buildings deformation monitoring. When extracting the buildings structurevibration signal, principal component analysis can be adopted to separate thecommon mode error and eliminate the inter station residual error with regionalsynchronous observation data. Then the method of wavelet analysis can be used toextract the buildings vibration signal, which would be verified from frequencydomain.