Research On The Key Models Of INS Aided BDS/GPS High Precision Kinematic Positioning

Integration of Global Navigation Satellite System(GNSS)and Inertial Navigation System(INS)can provide continuous,accurate and reliable positioning and navigation results,and it has been widely used for military purpose and civil use owing to their complementary characteristics.The development of rapid geoinformatics collection requires the support of high accuracy position information.The dissertation mainly focuses on INS aided and BeiDou Navigation Satellite System(BDS)/Global Positioning System(GPS)high precision positioning techniques,and the main researches include INS error stochastic modeling method,rapid INS initial alignment,BDS/GPS high precision integrated dynamic positioning,GNSS/INS integration model,adaptive interacting multiple model filter,INS aided cycle slip detection and repair,INS aided BDS/GPS single/dual frequency ambiguity resolution(AR)and INS aided precise point positioning(PPP).The work and contributions can be summarized as follows:(1)The INS navigation performance is directly affected by the efficiency and accuracy of initial alignment.A Rauch-Tung-Striebel(RTS)filter aided iterative initial alignment algorithm is proposed for quasi-stationary initial alignment.The RTS filter is used for attitude update after the forward initial alignment has been finished,the obtained attitude and sensor biases from the RTS filter are used as constraints in the iterative alignment process.The simulation test and field data test have verified that the proposed algorithm can improve alignment accuracy and shorten alignment time.(2)The frequent occurrence of cycle slips will degrade positioning performance,while the traditional cycle slip detection method,i.e.,observations combination,is mainly affected by code precision and ionospheric delay.An inertial aided tightly coupled(TC)cycle slip detection algorithm is proposed,the impact of INS positioning error on cycle slip detection has been analyzed,and cycle slip false alarm rate and cycle slip fixing rate are used as indicators to evaluate the algorithm.The results have shown that INS aiding can avoid the adverse effects from code noise and multipath,and it can effectively detect and fix the cycle slips during the occurrence of signal outages,the efficiency of INS aided cycle slip detection is dependent on the duration of outage.(3)The efficiency of GNSS ambiguity resolution is affected by the observation accuracy and observation environment.In dynamic scenarios,the ambiguity fixing efficiency is largely restricted,especially for single frequency case.An inertial aided ambiguity resolution algorithm is proposed by using INS predictions as constraints.The ambiguity search space can be reduced with inertial aiding,thus improves ambiguity fixing rate.In the constrained environments,the ambiguity fixing rate is restricted due to the existence of biased observations.An inertial aided partial ambiguity resolution(PAR)algorithm is proposed,the BDS/GPS integration model is constructed,and double differenced observations are decorrelated followed on,the ambiguity subset is selected based on maximum rounding rate.The fixed phase observations are then used to aid the other satellites ambiguity fixing.The results have shown that INS aided BDS/GPS system can obtained the best ambiguity fixing performance,the PAR scheme can improve the reliability of ambiguity fixing.(4)The ambiguity fixing efficiency will be degraded due to the occurrence of abnormal observations.In the urban dynamic environment,code multipath effect is obvious and the ambiguity fixing performance and positioning performance will be affected.In order to reduce the adverse effect from abnormal observations,a robust estimation algorithm with INS aiding is proposed;On the other hand,the accuracy of stochastic modeling will affect ambiguity fixing rate,while the a priori elevation based stochastic model is not accurate enough for modeling low elevation multipath.An adaptive on-line stochastic modeling strategy is proposed to improve ambiguity fixing performance.During BDS/GPS dual-frequency ambiguity resolution process,the stepwise ambiguity fixing strategy is adopted and the fading memory model is used to estimate the observation covariance.The results have shown that INS aided robust algorithm can improve ambiguity fixing rate,and the adaptive modeling strategy can further improve ambiguity fixing performance.(5)In long baseline conditions,the residual errors cannot be ignored after double differencing,thus degrade the ambiguity fixing performance.In order to improve the positioning reliability for long ranges,an INS aided ambiguity resolution algorithm with atmospheric constraints is proposed.The INS constraints and predicted atmospheric delays are used to aid BDS/GPS single epoch ambiguity resolution.The dual frequency data constraint is used to improve the reliability of ambiguity resolution,and the innovation based robust filter is constructed to detect the dynamic model fault or observation gross error.The results have shown that the ambiguity fixing rate can be improved with atmospheric constraints,and the dual frequency data constraint strategy can exclude false fixing,for long range baselines,the robust tightly coupled filter can obtain centimeter level positioning accuracy.(6)The optimal estimation of GNSS/INS relies on the modeling accuracy of system model and observation model,however,for the field navigation applications,the availability of a precisely known model is unrealistic.In order to improve the navigation performance,an adaptive interacting multiple model filter(AIMM)is proposed,the covariance matching strategy is applied to adapt process noise and measurement noise parameters.The UD based sequential updating strategy is used to improve processing efficiency.The results have shown that when adopting the code/Doppler/INS tightly coupled processing strategy,the GPS/BDS/INS integration system can obtain decimeter level positioning accuracy,and the proposed adaptive algorithm can improve the positioning accuracy.(7)Traditional PPP requires a relative long time to converge,thus limits its dynamic applications in some degree.In order to improve the PPP convergence rate,the INS aided tropospheric constraints model is constructed,and the PPP ambiguity is resolved based on the integer phase clock model,the adaptive estimation strategy is used to construct stochastic model.The results have shown that INS aided strategy can improve PPP convergence performance,and PPP ambiguity fixing mainly improves horizontal positioning accuracy,and tropospheric constraint can improve height accuracy and the adaptive filter can further improve positioning performance.