Bio-Inspired Geomagnetic Underwater Navigation by Multi-Agent Gradient Optimization

In the bio-inspired geomagnetic underwater navigation mission, the geomagnetic gradient can reflect the geomagnetic changes within a specific range, thus guiding agents to approach the target. However, the calculation errors of the gradient often challenge the accuracy geomagnetic trend. In this paper, we propose a multi-agent gradient optimization (MAGO) algorithm to improve the accuracy of geomagnetic target searching. Agents obtain gradient information at certain intervals at each time step. An optimization method is designed for optimizing geomagnetic gradient to calculate the heading angle. Further, a gradient comparison strategy is designed to overcome geomagnetic anomaly areas. The algorithm has good performance in the geomagnetic anomaly area. The experiments results verify the efficiency of the proposed algorithm.