【摘要】 为减小星三角形的冗余度,生成三角形数量最少、分布均匀性好的导航星三角形信息表,提出了一种以支持向量机为基础的导航星三角形选取优化算法.该算法通过一种以统计学习理论为基础的支持向量机寻求最优决策分类面,在由导航星表所生成的三角形中提取导航星三角形.实验结果表明,本算法所生成的导航星三角形表中的星三角形数量少,所需存储空间小,空间分布均匀性良好.更多还原

【Abstract】 The current guide star pick-up algorithms can not give a symmetrical distributing triangle, however, support vector machines based on statistical learning theory gives a new away to construct a new triangle data, with the least number and the most symmetrical distributing and the best performance, and reduce the triangle redundancy. A support vector machines triangle sorting algorithm is developed to seek the optimum decision separating hyperplane supporting vector machines based on statistical learning theory and get guide triangles from the ones ceated by guide catalog. The experiment results illustrated the algorithm is available.更多还原