Celestial Navigation and Positioning Method Based on Super-Large Field of View Star Sensors

As a crucial PNT component, star-sensor-based celestial navigation and positioning has been widely used in satellites, ships, aircraft, and other kinds of carriers. The larger the field of view of the star sensor, the more star points can be captured. Existing field of view of star sensors is generally not more than 20° × 20°, which limits the number of observed star points and creates difficulties for system calibration. Additionally, it is difficult to capture sufficient star points under cloudy weather conditions. This study proposes a celestial navigation and positioning method based on super-large field-of-view star-sensors, which can simultaneously image sky-wide objects, extending the effective field of view of star-sensors to 180° × 360°. The number of observable stars increases from tens to hundreds, facilitating system calibration and improving system usability. This study explores the corresponding algorithms of fisheye star image acquisition, star point extraction, star identification, and astronomical attitude determination algorithms. It is experimentally verified that the astronomical longitude error obtained using this method is −2.4 arcsecond, the astronomical latitude error is 4.7 arcsecond, and the corresponding position error is 153.2 m. If a higher resolution CCD is used, celestial navigation positioning with an accuracy of 30–50 m can be achieved.