Geomagnetic Surface Navigation Using Adaptive EKF

In need of accurate and covert positioning, a geomagnetic surface navigation system is investigated in the paper. The Earth anomaly magnetic field intensity is selected as the reference to bound the INS (inertial navigation system) errors. Using stochastic linearization technique, we have proposed an EKF (extended Kalman filtering) algorithm which is adaptive in two aspects: linearization region size determination and linearization noise variance estimation. Simulation results show that the proposed algorithm can achieve high positioning accuracy with moderate initial position errors and has great robustness with respect to the velocity errors, attitude errors and the geomagnetic map quality. The geomagnetic navigation system can be extended to air and underwater applications.