Gravity-assisted inertial navigation technology has been widely used in the underwater navigation communities. However, the traditional gravity matching algorithm has low efficiency and may cause out-of-domain mismatching, resulting in a significant increase in positioning errors, which affects the long-term high-precision navigation of underwater vehicles. To solve these problems, Iterative Double-cross Domain-center Transfer-matching Algorithm (ID2TM) is proposed with iterative double-cross line rough matching and local adaptive small-domain fine matching mechanism. The experimental results show that the proposed ID2TM algorithm has excellent matching performance in improving the underwater navigation positioning efficiency and matching reliability. Meanwhile, three tests in different gravity zones show that its worst matching accuracy and positioning efficiency are 93% and 84% higher than TERCOM on average, and the out-of-domain mismatching is completely avoided, which further verified that

The matching low-efficiency and mismatching of the domain-based sequence matching are two main problems affecting the underwater navigation ability. Compared with the in-domain mismatching, the out-of-domain mismatching occurs outside the matching domain and its positioning error is larger, which results in more serious navigation effects. To address these problems, a cross-line adaptive domain matching (CADM) algorithm is proposed to improve the positioning efficiency and out-of-domain matching capacity of underwater navigation. Different from the canonical sequence correlation matching algorithm, the proposed method realizes the domain-center adaptive screening by using a cross-line pre-matching trick, and obtains a better repositioning for the vehicle’s real position by employing an adaptive domain re-matching mechanism. Simulation results show that the proposed method can effectively improve the positioning efficiency and out-of-domain matching performance of underwater gravity matching navigation. In two track tests, the matching efficiency is averagely improved by 82%, the number of the out-of-domain mismatches are reduced by 94.87% and 97.22%, and the improvement ratios of the out-of-domain matching accuracy indexes are almost over 86%, which adequately verify the outstanding matching capacity of the proposed method. And it can provide a new choice for the underwater autonomous passive navigation in the future.