Data Fusion Technology In Integrated Navigation Systems

Taking SINS/GPS/TCM2 integrated navigation system as the object, this paper gives a study on the federated filter data fusion technology, which provides an effective way to realize the best performance of SINS/GPS/EC (Electronic Compass) integrated navigation system in theory and practice.The paper briefly introduces the federated filter theory. In this paper, the equivalence between centralized and federated filter is proved. Four typical structures are introduced, and their characteristics of fault tolerance are analyzed.Based on the analyses of every subsystem, their error models and the methods how to integrate them are set up. And then, a centralized filter is designed. Simulation results show that this centralized filter is effective to be applied to SINS/GPS/ EC (Electronic Compass) integrated navigation system.According to the discussion of federated filter theory and SINS/GPS/EC (Electronic Compass) integrated navigation system, a federated filter is designed and simulated. Based on the demands of real-time and rapidity, two simplified federated filter are introduced. Simulation results show the simplified algorithm has the advantages of less computing and data-communicating amount.The effects of information-sharing coefficient to the federated filter are researched in this paper. In order to get satisfactory fusion accuracy and the best performance of the subsystem, the information sharing principle is that information-sharing coefficient should be inverse proportion to the precision of the subsystem.The problem of designing fault-tolerant integrated navigation systems is also discussed. The paper introduces the theory of two kindsof the chi-square test and proposes the scheme of fault detection and isolation that is suitable to engineering based on the simulation results of the SINS/GPS/EC (Electronic Compass) integrated navigation system.