Electromagnetic Fusion Underwater Positioning Technology Based on ElasticNet Regression Method

This paper presents a novel underwater positioning method based on the electric field and magnetic field data fusion, which belongs to non-acoustic positioning technology. The related development results of this technology can be applied to underwater detection and rescue, the location of underwater unidentified vehicles, and the tracking records of marine aquatic products. In order to improve the processing speed, we propose an optimization method based on the ElasticNet Regression (E-R) method. The optimization principle is to configure the distribution function by introducing the lasso convergence term. The structure and reasonable configuration of the lasso convergence term coefficient can improve the convergence speed of the algorithm. It is found that the electromagnetic field data is simulated and collected by CST simulation software, and the electromagnetic field data is input into the positioning algorithm, and a large number of random experiments are completed. Through the chart comparison, the convergence speed and accuracy of the electromagnetic fusion positioning algorithm are analyzed.