Matching area selection for arctic gravity matching navigation based on adaptive all-field extended extremum algorithm

Suitable and effective matching area selection is crucial for gravity matching-aided navigation. In this paper, an all-field extended extremum algorithm based on an adaptive threshold (AT-AEE) is proposed for matching area selection in the Arctic Sea. The gradient data is obtained by using the convolution of gravity reference graph data and all-field extended extremum parameters. Then, the adaptive threshold method was employed to determine the optimal gradient threshold based on gravity anomaly data across various test areas. Data points with gradients exceeding the specified threshold are identified as local candidate points for matching areas. The test areas containing a certain proportion of local candidate points are designated as the suitable matching areas. Nine test areas in the Arctic Sea with different gravity change characteristics were chosen for simulation experiments to verify the performance of the proposed algorithm. Simulation experiments showed that superior navigation positioning results could be obtained in the matching areas selected by the AT-AEE algorithm. Compared to traditional algorithm, the matching areas derived from the AT-AEE algorithm performed with a better consistency in the gravity matching navigation results. In suitable matching areas with the proportion of local candidate points reaching 70%, the average positioning errors could be reduced to less than 1.5 n miles. Suitable and effective matching area selection is crucial for gravity matching aided navigation. In this paper, an all-field extended extremum algorithm based on adaptive threshold (AT-AEE) is proposed for matching areas selection in the Arctic Sea. Compared to traditional algorithm, the matching areas derived from the AT-AEE algorithm performed with a better consistency in the gravity matching navigation results. In suitable matching areas with the proportion of local candidate points reaching 70%, the average positioning errors could be reduced to less than 1.5 n miles. image