Improving the Out-of-Domain Matching Reliability and Positioning Accuracy of Underwater Gravity Matching Navigation Based on a Novel Cyclic Boundary Semisquare-Domain Researching Method

Abstract

A gravity matching algorithm is a crucial component of a vehicle’s underwater navigation system. This article studies the improvement in the out-of-domain matching reliability and positioning accuracy of underwater navigation. To achieve these goals, a novel cyclic boundary semisquare-domain researching (CBSR) method is proposed. Two times the inertial error is first employed to span an initial small-square domain and find an initial optimal matching position, which results in controlling the matching efficiency. If this optimal position is located on the domain boundary, the cyclic boundary semisquare domain rematching mechanism is triggered. The cyclic boundary semisquare domain is iteratively generated to perform repositioning and obtain a better matching position until it falls into the interior of the matching domain. The final optimal matching position is used to calibrate the sensor’s parameters and aid the navigation of the underwater vehicle. Experimental results prove that the proposed CBSR method has outstanding positioning capacity for underwater gravity matching navigation, and higher matching reliability and lower mismatching probability of the vehicle’s out-of-domain positioning. For excellent and good suitable tracks, the number of out-of-domain mismatches of the CBSR method, compared with TERCOM, are reduced by 97.73% and 86.05%, respectively, while the out-of-domain average matching accuracies are all less than one grid resolution and are improved by 85.23% and 81.07%, indicating the effectiveness and feasibility of the proposed CBSR model for improving the matching reliability and positioning accuracy of out-of-domain underwater navigation.