.png){kind=icon}
.png){kind=icon}
Return
A Star Identification Graph Algorithm Based on Angular Distance Matching Score Transfer
.png){kind=icon}
PRE
AI
DOI：10.1109/JSEN.2024.3350089.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The successful identification of stars is afundamental prerequisite for satellite attitude determinationby star sensors. Conventional star identification algorithmstypically construct specific subgraphs using a set of brightstars or generate patterns based on angular distancesbetween stars and their neighboring stars. However, thesemethods often fail when insufficient detectable stars or brightnoncooperative space objects are in the field of view (FOV).Seldom studies use angular distance matching among allstars for direct identification as it is complex to calculate andstore. To solve the problem, we proposed a graph algorithmbased on angular distance matching score transfer, whichutilizes a graph data structure, where nodes store angulardistance matching scores between stars and transfer scoresthrough graph edge weights, overcoming the limitations ofincomplete utilization of angular distance matching results.We also established a math model for edge weights basedon the probabilities of star pair occurrences. Simulation testsand night sky image experiments demonstrate the robustness of this algorithm against position errors, brightnesserrors, and fake stars. By equally using information from all sensor stars, even in cases with multiple missing stars,numerous fake stars, and extremely bright interfering stars, the identification rate remains consistently above 99.53%.This algorithm represents a highly robust star identification method and lays the foundation for future research inintelligent sensing of space environmental objects using star sensors.
Keywords:
Angular distance matching
graph representation algorithm
star identification
star sensor