| description abstract | In the last decade, the demand for accurate land-vehicle navigation (LVN) in several applications has grown rapidly. In this context, the idea of integrating multisensor navigation systems was implemented. For LVN, the most efficient multisensor configuration is the system integrating an inertial navigation system (INS) and a global positioning system (GPS), where the GPS is used for providing position and velocity and the INS for providing orientation. The optimal estimation of the system errors is performed through a Kalman filter (KF). Unfortunately, a major problem occurs in all INS/GPS LVN applications that is caused by the frequent GPS signal blockages. In these cases, navigation is provided by the INS until satellite signals are reacquired. During such periods, navigation errors increase rapidly with time due to the time-dependent INS error behavior. For accurate positioning in these cases, some approaches, known as bridging algorithms, should be used to estimate improved navigation information. In this paper, the main objective is to improve the accuracy of the obtained navigation parameters during periods of GPS signal outages using different bridging methods. As a first step, three different KF approaches will be used, including the linearized, extended, and unscented KF algorithms for the INS/GPS integration. Two land-vehicle kinematic data sets with different-quality INSs are used with several induced GPS outages, and then two bridging approaches are implemented. The first method is to apply different backward smoothing algorithms postmission that are associated with the different used KF approaches. The second bridging method is a near real-time approach based on developing an INS error model to be applied only during GPS signal blockages. After applying each bridging method, the results showed remarkable improvement of position errors regardless of the KF used. | |
