Comparison of State and Error State INS Coupling Filter Based on Real Flight Test Data

Ulf Bestmann, Meiko Steen, Martin Becker, Andreas Sasse, and Peter Hecker

Abstract:	In the past decades filtering of INS data together with aiding information especially GNSS has been an innovative moving science. In the last years much work was spend on innovative filter concepts like unscented kalman filters or particle filters. There is also a huge interest in the kind of aiding values used and the way they are fet into the filter. Aiding values can be for example precalculated values like position and velocity of a GNSS or among others as raw range, doppler and carrier phase measurements from a satellite. There are also different concepts in which way the filtered information can be feed back to the process and to the hardware. Last but not least there are numerous different filter models which observe a different number of states with different assumptions made on sensor behavior and linearisation. A basic aspect of INS filtering is the question, wether the filter is used to observe the errors of the states in the INS and the strapdown process (indirect observer) or if the filter comprised the strapdown process and does not observe the errors of the states but instead the states itself (direct observer). Some advantages and disadvantages of the direct and indirect observer are obvious. The error state filter has the advantage that it observes the most of the time small errors of the states, which, if they are fed back to the strapdown calculation, stay small during the whole calculation. This makes linearized models behave well in the error state filters. Especially facing the Euler angles, observing error states has the advantage, that it is not necessary to use quaternions in the filter but only in the strapdown calculation. This reduces the states by one. On the opposite the state filter concept has the advantage, that there is no need to feed back the observed states to an external computation, because the states are used in the filter in a filter dependent and mathematical ideal manner. This paper shows how the different filter architectures perform with real world data from flight tests.
Published in:	Proceedings of the 20th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS 2007) September 25 - 28, 2007 Fort Worth Convention Center Fort Worth, TX
Pages:	2611 - 2618
Cite this article:	Bestmann, Ulf, Steen, Meiko, Becker, Martin, Sasse, Andreas, Hecker, Peter, "Comparison of State and Error State INS Coupling Filter Based on Real Flight Test Data," Proceedings of the 20th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS 2007), Fort Worth, TX, September 2007, pp. 2611-2618.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In