Vorticity Fields From Polarization Neighborhood Variations to Improve Solar Tracking for Autonomous Integrated Navigation

Polarization navigation utilizes atmospheric polarization patterns for orientation, offering a reliable alternative in Global Navigation Satellite System (GNSS)-denied environments. Precise solar tracking is essential in polarization navigation, as the solar vector connects polarization patterns to navigation information. However, existing methods face two key limitations: weak geometric perpendicular constraints, leading to underdetermined solar vector estimation, and insufficient fusion of polarization features, reducing accuracy. To address these challenges, this paper proposes the polarization vorticity field (POL-VF) method, which integrates the spatial variation of the angle of polarization (AoP) and the degree of polarization (DoP) to establish collinearity constraints, improving solar tracking accuracy. Dynamic trials validate its enhanced performance in autonomous integrated navigation, achieving a heading Root Mean Square Error (RMSE) of 5.18° over a 40km ground vehicle test and 5.6° at an 8m depth in underwater experiments.