Locating and tracking of underwater sphere target based on active electrosense

Underwater active electrosense serves as a valuable complement to conventional underwater target detection technologies, such as acoustic and optics, by offering advantages in short-range and murky environment. However, there is a need to enhance the precision of target location and enrich the research on tracking moving target. This paper delves in to the investigation of locating and tracking underwater sphere target using active electrosense. To improve location accuracy, an approach is proposed to generate a global adaptive model correction coefficient through calibration and general regression neural network (GRNN) fitting. A detection system and sensing array with vertically arranged transmitting and receiving electrodes are designed for experimental research, enabling high-resolution target locating and tracking without the need of any active movement of the array. In the scenario where the target is metallic sphere with a diameter of 50 mm, the results of location experiment indicate that the least square (LS) algorithm achieved 24.7% higher location performance than the multiple signal classification (MUSIC) algorithm under an accurate uniform model correction coefficient, and the location error of the LS algorithm was further reduced by 33.2% using the global adaptive model correction coefficient generated by proposed approach. Based upon these findings, tracking functionality based on discrete location method is achieved under various trajectories and velocities.