基于支持向量机的导航星选取算法研究

在星敏感器导航星表的建立过程中由于恒星的数量太多, 往往要进行筛选, 通常这种选择是一种基于枚举的大量反复的提取过程, 复杂费时而结果往往并不是最优的。而基于统计学习理论( SLT) 的支持向量机( SVM) 方法正好克服了这方面的不足。SLT 理论和SVM 方法为导航星选取过程的简化和结果的最优性的获得提供了新的途径。讨论了支持向量机在导航星选取优化中进行应用的分类算法, 构建了导航星分类器, 并以导航星的选取为例进行了试验论证。试验表明: 基于SVM 的导航星分类器对简化导航星的筛选过程优化导航星表的

Application Research on Support Vector Machines in Optimal Selection of Guide Star

The optimal selection of the guide star is important and necessary to generate the on-board guide star catalog since the number of the stars is too large to be fitted with the star tracker. The existing general selection procedures of the guide star, based on the large number of enumeration and many times iteration, are usually complicated and time-consuming, and the result generated by these selection methods is not always optimal. The Support Vector Machines ( SVM) based on the Statistical Learning Theory ( SLT) can solve this problem. SVM and SLT provide us a new approach to develop the simplified procedure of guide star automatic selection. The classification selection algorithm of the guide star using the SVM is discussed in this paper and the classifier of the guide star is obtained. Experiments on selection of the guide star are conducted and the results demonstrate that the SVM-based classifier of guide star has high adaptability for the automatic selection of the guide star.