摘要
为提高大视场星敏感器的星图识别速度和识别成功率,提出一种基于蚁群聚类算法的快速星图识别方法,该方法首先利用蚁群聚类算法对星点集合进行快速聚类分析;然后选择最优类并以其中每个星点为圆心,以一定角距为半径画圆,将圆内所有星点构成集合;再将每个集合的星点两两求取角距,利用蚁群算法对路径快速寻优的优点,完成集合的路径优化;最后利用优化结果同导航星库中已有的优化数据相匹配,以实现星图的快速匹配与识别;实验结果表明,与现有识别方法相比,该方法具有高的识别率,良好的实时性和鲁棒性,且所需导航星库的容量小。
In order to increase star map identification speed and successful rate for large field of view, a rapid star map identification method based on ant colony clustering algorithm is presented. This method adopts ant colony clustering algorithm to quickly carry out clustering analyze for star-points set, chooses the optimal class and draws one circle by the center of circle being every star point in this class and the radius being special angular distance, and then composes a set including all star points in circle. Then angular distance of two star points is solved in each star-points set, and completes the path optimization of sets by ant colony algorithm. At last, the rapid match and recognize is be done by comparing the result of optimization with the corresponding data in guidance star database. The experimental results show that, compared with existing identification methods, this method has a high recognition rate, good real-time performance and robustness, and it only needs a small content database.
出处
《宇航学报》
EI
CAS
CSCD
北大核心
2008年第6期1814-1818,1870,共6页
Journal of Astronautics
基金
国家自然科学基金项目(60574086)
关键词
蚁群算法
聚类分析
星图识别
导航星库
Ant colony algorithm
Clustering analysis
Star map identification
Guidance star database