State-Space Measurement Update for GNSS/INS Integrated Navigation

Abstract

This paper theoretically derives the equivalence conditions for the loosely and tightly coupled GNSS/INS integration algorithms. Firstly, the equivalence is proved when using single epoch GNSS measurements, which means the GNSS processor provides standalone solution. Then, the equivalence proof is further extended for the filtering solutions, which are usually applied for differential GNSS and precise point positioning. Based on these, different state and measurement models for GNSS/INS integration navigation are summarized, and natural differences among these models are discussed. This indicates that once the same measurement and predict information are used, the integration would be equivalent no matter what kind of coupling schemes are used. A flight dataset with GNSS and tactical IMU data is used to evaluate the equivalence and discrepancies among four different measurement models, and the results confirm the theoretical derivations.