Optimized Magnetic Gradient Tensor Array Algorithm for Enhanced Positioning Accuracy of Underwater Weak Magnetic Targets

This research utilized an 8-fluxgate array, based on the magnetic gradient tensor (MGT) method, to perform positioning trials within a controlled aquatic environment. The experiment featured five designated locations and targeted marginal magnetic signals of approximately 2–3 nT, which are resolvable by the sensors. The standard deviation of the positioning outcomes ranged from 0.0724 to 0.0764, while the root mean square error (RMSE) was confined to 0.0921–0.1468. At a detection distance of about 6 m, the discrepancies in the x- and y-axes were comparatively minor, whereas the z-axis exhibited a more significant deviation in positioning. By comparing with scalar triangulation and ranging (STAR), which is also based on the MGT algorithm, and incorporating the averaging method from their approach, we have optimized the original MGT algorithm. This enhancement has led to improved positioning accuracy, with the RMSE reduced to a range of 0.0219–0.0734, resulting in superior localization performance.