Filtering Algorithms And Reliability Analysis For GNSS/INS Integrated Navigation Systems

As a non-autonomous navigation technique,the GNSS(Global Navigation Satellite System)is able to provide rapid,continuous and precise information of navigation and positioning.However,the GNSS is affected significantly by signal quality of the satellites.INS(Inertial Navigation System)is an autonomous navigation technique without the problem of signals,but it takes a long time in the initial alignment and the errors accumulate quickly with time.Properties of GNSS and INS are complementary and integration of them becomes inevitable.The GNSS/INS integrated navigation systems have been broadly applied in military and people's production and life.Along with the demand of location service,requirements of accuracy,stability and reliability for the integrated navigation and positioning are increasing.Filtering algorithms and reliabilities of the GNSS/INS integrated navigation systems are studied and analyzed in this paper.The main contents are listed below:error modeling of the strapdown INS(SINS),modeling and performance analysis of the GNSS/INS integrated navigation systems,modeling and applications of the robust interval Kalman filter and the adaptive H?filter,adaptive estimation of the multiple fading factors and its application in the GNSS/INS integrated navigation systems,and the reliability analysis of integrated navigation systems et al.The main contributions obtained in this paper are as follows:(1)Models of the loosely-coupled and tightly-coupled GNSS/INS navigation systems are derived and constructed,and the corresponding filtering are processed with self-designed softwares.With the data collected by the self-designed GNSS/INS integrated navigation systems under actual conditions,contrastive analysis of the precision and performance is performed between the loosely-coupled and tightly-coupled GNSS/INS navigation systems.(2)For the GNSS/INS integrated navigation systems,two problems need to be addressed properly:modeling and parameter estimation.The former need to establish a proper error model and the latter requires appropriate parameter estimation algorithm and measurement information.Aiming at problems of the dynamic model inaccuracy and the abnormal observations,the reduced interval Kalman filter is constructed to relief the computation burden,then the robust interval Kalman filter is proposed,and influences of model errors and abnormal measurements are controlled effectively with the proposed algorithm;aiming at problems of the dynamic model inaccuracy and the unknown interferences,the adaptive H-infinity filtering algorithm is proposed with the double-factor robust estimation method;it is difficult to estimate the existing multiple fading factors,thus,a new adaptive estimation method of the multiple fading factors is developed based on the single optimal fading factor,and this provides beneficial reference for the practical work.(3)Aiming at reliability of the GNSS/INS integrated navigation systems,the theory of extended receiver autonomous integrity monitoring is introduced and applied.The statistics obey?~2 distribution and w hypotheses test based on the least squares residual are adopted to detect and identify the malfunctions.Then,a new robust ERAIM algorithm is proposed which is able to provide reliable statistics for the integrity monitoring,thereby ensuring the filtering reliability and reducing the probability of false alarm for the GNSS/INS integrated navigation systems.