Intelligent Path Planning for Navigation in Complex Marine Geomagnetic Environments

Geomagnetic navigation provides significant promise in GNSS-denied environments, owing to its passive characteristics and global availability. However, its accuracy heavily depends on geomagnetic features, with substantial degradation in geomagnetically sparse regions. In contrast to conventional path planning, this paper proposes a navigation-oriented framework centered on establishing a novel metric for geomagnetic suitability. This metric, termed complexity, quantitatively assesses the geomagnetic information richness of each grid, serving as a foundational map. Building upon this, we develop a geomagnetic-complexity-aware deep reinforcement learning (DRL) path planning method that optimizes paths toward high-complexity regions, achieving an optimal trade-off between path length and navigation ability. Furthermore, we integrate this metric into classical path planning algorithms for comparative analysis. Experimental results validate that the proposed framework significantly enhances the navigation performance of autonomous underwater vehicles (AUVs) in complex marine geomagnetic environments, enabling adaptive path planning with improved positioning accuracy and operational efficiency.