A Cascade Gravity Matching Algorithm With Large Initial Position Error

Gravity-aided navigation is an effective passive navigation method, which has a promising application in underwater autonomous vehicles. In order to enhance the real-time performance of the gravity matching algorithm, improve the positioning accuracy of the inertial navigation system (INS), reduce the influence of unknown initial position error, and strengthen the applicability of the gravity-aided navigation system. This article proposes a cascade algorithm based on the maximum correlation method and the extended Kalman filter (EKF) for the gravity-aided navigation system. This algorithm obtains the matching result of the maximum correlation method by the mean square difference (MSD) first. Then the result sequence of MSD and the gravity anomaly variation are introduced into the observation equation of the EKF to correct the position of INS. The effect of the initial position error can be reduced by the MSD method, while the EKF algorithm can improve the real-time performance of gravity matching. The postprocessing simulations of the actual ocean experiment show that the gravity-aided navigation cascade algorithm proposed in this article has good precision. And this algorithm also improves the stability and the applicability of the gravity matching algorithm.