摘要
导航恒星提取一般采用星等过滤方法MFM(Magnitude Filtering Method)。但是MFM方法存在两个明显的缺陷:若星等阈值太高,导航星表冗余度高;反之,导航星表出现视场(FOV)空洞。支持向量机SVM(Sup-port Vector Machine)作为一种可训练的机器学习方法,依靠小样本学习后的模型参数进行导航星提取,可以得到分布均匀且恒星数量大为减少的导航星表。利用SAO星表进行了实验,并对导航星表内恒星的分布情况作了统计。实验证明,SVM作为导航星提取算法具有很好的应用前景。
The method of constructing navigation star catalogue is always based on Magnitude Filtering Method ( MFM). But it does not work well because of two typical disadvantages. On one hand it will extract so many stars that there is redundancy in the catalogue. And on the other hand it will generate "hole" in some area of celestial sphere. In this article, Support Vector Machine(SVM) is introduced into extracting navigation stars from basic catalogne. After using the new method on SAO catalogue, it is proved that taking SVM as the method of extracting navigation-stars has good prospection.
出处
《测绘科学技术学报》
北大核心
2007年第3期196-199,共4页
Journal of Geomatics Science and Technology
基金
国家自然科学基金资助项目(40571131)
关键词
导航星表
卫星姿态
机器学习
SVM
navigation-star catalogue
satellite attitude
Machine-learning
SVM