| Global navigation satellite system(GNSS),especially global positioning system(GPS),which can provide real-time accuracy positioning information,is widely used in many Intelligent Transport Systems(ITS)applications.In urban environments,GPS signals are often reflected,blocked or diffracted by buildings,which resulting in multipath effects and non-line-of-sight reception(NLOS).These effects also cause range errors that degrade the accuracy of GPS positioning,which will hinder the wide applicaiotns of ITS.Therefore,the research on the GNSS positioning accuracy improvement in urban multipath environment is of great importance forthe national and international ITS development.This theses first introduces the principle of GPS positioning,the error sources and corresponding error mitigation methods.Afterwards,the existing GPS multipath mitigation methods are reviewed.A new GBDT based pseudorange error prediction model is proposed by using carrier-to-noise ratio(C/N0),pseudorange residuals and satellite elevation angle as the input variables.With the predicted pseudorange errors,two variations of the algorithm are proposed to improve the positioning accurcy,including the pseudorange error correction and the multipath/NLOS exclusion or correction.The proposed method is tested in two different scenarios,and the performance is evaluated from the pseudorange prediction accuracy,signal reception type classification accuracy and positioning accuracy.The test results show that the proposed method can effectively improve the positioning accuracy in urban multipath environment,and the positioning accuracy based on pseudorange error correction is superior than that based on multipath/NLOS elimination or correction. |