Fuzzy Logic vs. Kalman Filtering Techniques for Marine GPS/INS Integration When Centralized Filtering is not Possible

Griffith McRee

Abstract:	When third party integrators have access only to GPS/INS information which has already been KalmanFiltered, they are faced with a dilemma. The use of a traditional centralized Kalman Filter in such situations violates the assumption that input noise is white. Currently, they are left with only two options: the Cascaded Filter approach or the Federated Filter. Both options have drawbacks, but improvements are being attempted. The RF has served the navigation engineer well, but this should not eliminate the search for alternate mathematical techniques when the application of the KF becomes forced. Fuzzy logic is a mathematical technique that allows human knowledge to be encoded numerically. Fuzzy logic controllers have been shown to be robust and accurate. The most famous application to date is the control of the Sendai subway system in Japan. It is the opinion of the authors that the fuzzy logic technique should be examined for navigation systems integration when KF techniques are being used in situations that cannot be theoretically justified. The objective of this work is to design and code a fuzzy logic based estimator and to compare it with the more traditional KF using computer simulations. The suite of marine navigation equipment to be integrated consists of two GPS receivers (one PPS and one SPS), an Omega receiver, an E.M. Log, and an INS. Overall performance as well as FDI capabilities will be examined. The availability of continuous global GPS coverage changes the integration strategy of marine INS/ Radionavigation systems. Previous methods of INS up- date used less accurate Radionavigation systems and updated less frequently. Marine navigation system inte- grators are now confronted with the fact that more infor- mation exists than current algorithms can fully assimilate. One current solution is to reduce the rate of information exchange between GPS and INS. The difficulties encountered by third party integrators will increase in number and complexity as more and a greater variety of receivers enter the market place. Furthermore, it is anticipated that as manufacturers strive to succeed in a competitive market place, they will continue to regard as proprietary the information required by third party inte- grators to build a Centralized KF. Thus, the need for third party integration techniques such as fuzzy logic will become more important. This paper will also include an introduction to fuzzy logic for those navigation engineers who are unfamiliar with the technique. It would be significant for navigation systems integration if new mathematical techniques that will aug- ment RF techniques could be found.
Published in:	Proceedings of the 6th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GPS 1993) September 22 - 24, 1993 Salt Palace Convention Center Salt Lake City, UT
Pages:	215 - 215
Cite this article:	McRee, Griffith, "Fuzzy Logic vs. Kalman Filtering Techniques for Marine GPS/INS Integration When Centralized Filtering is not Possible," Proceedings of the 6th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GPS 1993), Salt Lake City, UT, September 1993, pp. 215-215.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In