INS error model improvement for enhanced INS/GPS navigation during GPS signal blockage periods

Abstract

Any integrated navigation system that consists of an Inertial Navigation System (INS) and a Global Positioning System (GPS) suffers sometimes from a major problem: the existence of frequent GPS signal blockages (GPS outage periods). In this case the INS is used for navigation but navigation errors increase rapidly with time. One of the main factors that affect such obtained errors is how the INS sensor errors are modeled. For most IMUs, a 1st order Gauss-Markov (GM) process is usually used. In this paper, it will be shown that this process is not always fitted with the behaviour of inertial sensor errors. Instead, Autoregressive (AR) processes of orders higher than one are suggested. Numerical analyses are performed to compare the accuracy of GM and AR parameters that are obtained from experimental data. Using GPS and a tactical-grade IMU, a kinematic INS/GPS data set is used for testing the performance of GM and AR models during GPS outages. The results showed a considerable reduction of position errors in these periods when implementing AR processes instead of GM ones.