Design & Research In Automatic Fault Tolerance Integrated Navigation System

It is an effective way in integrated navigation by using Strapdown Inertial Navigation System(SINS),GPS and Doppler Velocity Log(DVL). This paper studies SINS/GPS/DVL integrated navigation whose emphases is Kalman Filter, and failure detection.This paper prepares two traditional Kalman filter which estimate the position and velocity of vehicle. Due to the model's inexactitude of traditional Kalman filter, we discuss adaptive filtering and BP neural network which are the improvement of traditional Kalman filter.For the assurance of observational data's dependability, we must detect and isolate the failure during the process of Kalman filter, and failure detection is the emphase.At last, the paper validates the theory of Kalman filter and fault tolerance by the soft of automatic failure detection integrated navigation.