A Variable Proportion Adaptive Federal Kalman Filter for INS/ESGM/GPS/DVL Integrated Navigation System

An integrated navigation system with synchronized space and time is constructed in this paper, which acquires information primarily from INS and ESGM, and GPS and DVL information as secondary sources. The error models of subsystems are set up respectively and the parameters matrixes of the state equation and measurement equation are fractional step discretization. A novel variable proportion adaptive federal Kalman filer algorithm is proposed for information fusion. The weights of filters are adjusted according to the work status of navigation sensors and validity of navigation information. Simulations are performed and the results show that high precision, stability and fault-tolerance can be obtained.