Research On Key Technologies Of BDS/GNSS Uncombined Precision Point Positioning

With GPS and GLONASS systems operating in full constellation and continuous updating,and Galileo system accelerating global layout,and Beidou global satellite navigation system initially having the ability to provide services to the world by the end of 2018,BDS/GNSS PPP(Precise Point Positioning)has become one of the hotspots in the field of GNSS research.More systems bring PPP more available satellites and better satellite constellation space geometry layout,greatly shorten the convergence required by PPP and improve the positioning accuracy,which have a broad application prospect.Compared with the traditional ionosphere-free model,the uncombined PPP model retains more useful parameters,reduces the observation noise,and is more conducive to the application expansion of multi-system and multi-frequency mode,which is valuable and useful.Based on these,the paper focuses on BDS/GNSS uncombined PPP,and conducts in-depth research from PPP cycle slip detection and repair,BDS uncombined PPP model,to BDS/GNSS multi-system uncombined PPP and their inter-system biases.The main research contents and innovative work are as follows:(1)The application of the second-order time-difference phase ionospheric residual(STPIR)and front-to-back sliding window filtering method which is based on wide lane combination to detect and repair the cycle slip is discussed.The experimental results show that the algorithm can effectively detect and repair the simulated cycle slips even under the condition of large magnetic storm,and there are no insensitivity cycle slips.(2)In order to detect and repair BDS2 and BDS3 triple-frequency cycle slips,we construct two geometric-free phase combinations and one ionospheric-free pseudorange-phase combination to form three linearly independent detection measurements of cycle slips.In order to tackle the ill-conditioned problem of the normal matrix when we repair cycle slips,a new method for repairing BDS2 and BDS3 triple-frequency cycle slips based on extended ridge estimation is proposed.Due to the limit of range and magnitude of the ridge parameter k,the traditional method of selecting ridge parameter does not work.Aiming at this difficulty,ridge parameter search method based on the principle of minimizing the sum of squares of residuals is proposed,which is highly effective and accurate.The sum of squares of the two groups of residuals is set as the judgment index of the correctness of cycle slip repair,which guarantees the reliability of cycle slip repair.Finally,the simulation results show that the algorithm is reliable effective in repairing BDS2/BDS3 triple-frequency cycle slips.(3)There is serious long period multipath effect(LPME)caused by long period multipath effect in the BDS GEO satellites,and the magnitude of which can reach meter level.The impact of LPME is slowing down the convergence and reducing the reliability of positioning.In order to solve the difficulty,the uncombined PPP model considering the LPME of BDS GEO satellites is studied and derived,in which the parameter LPME is regarded as a constant to estimate within one day,which increases the model stability.And experimental results show that,compared with the schemes of ignoring GEO satellites and reducing the weight of GEO satellites,BDS uncombined PPP which estimates LPME of GEO satellites can effectively reduce the pseudo-range residual of PPP,speed up the convergence and improve the accuracy slightly.(4)We have studied the BDS uncombined PPP model that estimates receiver DCB(Differential Code Bias,DCB),and deduced the function model.Then,we evaluate the influence on PPP when white noise model,random walk model and constant model are used to estimate the receiver DCB respectively.And the characteristics of change of the receiver DCB sequences estimated by the three models are studied.The experimental results show that,compared with the traditional uncombined PPP,using the white noise model as the parameter estimation model of receiver DCB can improve the positioning accuracy of PPP slightly,but the improvement of convergence seems slim;when using the other two groups of parameter estimation models to estimate the receiver DCB,it takes no advantages in positioning accuracy and convergence of PPP compared with traditional uncombined PPP;The receiver DCB estimated by constant model and random walk model are close to each other,and after the filter converges,they both show great intra-day and inter-day stability,while the receiver DCB sequence estimated by the white noise model is not stable in one day.(5)We have studied the BDS/GLONASS uncombined PPP.The fusion model considering pseudorange long period multipath effect of BDS GEO satellites and the pseudo-range inter-frequency biases(IFB)of GLONASS satellites is derived,and the positioning performance between BDS uncombined PPP and BDS/GLONASS uncombined PPP is compared.Experimental results show that BDS/GLONASS uncombined PPP takes significant advantages in BDS uncombined PPP in the number of available satellites,positioning accuracy and convergence.(6)We have deduced uncombined PPP function model of GPS,BDS,GLONASS and Galileo systems and the expressions of three groups of ISB(inter system bias)parameters,in which the GLONASS pseudo-range IFB is regarded as a constant in one day to estimate,which simplifies the model structure.Then,the positioning performance of four system uncombined PPP is discussed,and the characteristics of the ISB are analyzed in detail.Experimental results show that: 1)uncombined PPP of four systems can provide great positioning accuracy and convergence,and can effectively guarantee the reliability of ISB estimation;2)the intra-day variation of the ISB sequences are relatively stable,but the inter-day variation of them are pretty intense,and there is a clear correlation between ISB parameters and receiver types.