Troposphere Delay Regional Fitting Model Based On CORS Network In PPP

Precise point position (PPP) technology takes advantage of GPS post-processing preciseephemeris and precise satellite clock error, with the phase data acquisitioned by a singledual-frequency GPS receiver as the main measurements for single point position, it’sprecision can reach cm level.With the improvement of precise point position technology, some error sources in thePPP data processing take the people’s attention. Among them, the satellite clock error,ionosphere delay and troposphere delay is there key error sources in GPS precise pointposition. So to get high sampling interval precision clock difference products, establish theionosphere second order correction model, to obtain high precision of troposphere delaycorrection is a hot spot in current research.I introduced in this paper the latest progress in the satellite clock error and ionospheredelay, mainly for the PPP theory and implementation method of troposphere delay correctionin the depth study and research. Troposphere delay is an important error in precise pointposition, which estimation precision directly influences the ultimate position precision. Inprecise point position, generally, we should first estimate the troposphere delay model, thenthe remaining residuals for parameter estimation. With the establishment of GPScontinuously operating reference station(CORS), the troposphere information based on theCORS network can provide high accuracy estimator of troposphere delay correction, throughthe data we can set up the region of troposphere delay correction fitting model, and then tointerpolation fitting model for rover of troposphere delay correction information. This methodcan not only improve the location accuracy of PPP, and can increase the convergence rate ofthe PPP.At present commonly used method of fitting are a curve fitting, plane fitting of threeparameters, quadric surface fitting, the polyhedral function fitting and so on. However theabove fitting method only considers the fitting error of the dependent variable in the model,without considering the error of the independent variable, is obviously unreasonable. Basedon the above problems, this paper puts forward to the two at the same time considering the dependent variable and independent variable error fitting model, three parameter fittingmodel and a yuan of linear fitting model based on the overall square respectively. Throughsetting up mathematical model of the observational data, because of the complexity of theactual problem and observation data inevitably contains errors, such error leads to the modelis often inaccurate, so need to compensate to the model. This paper had a compensation to acurve fitting by the least square collocation method, established a configuration based on theleast square curve fitting model, and verified the above-mentioned three kinds of new modelby the program made by myself.