基于核密度估计的极短弧定轨的分布估计方法  

Estimation of Distribution Algorithm for Initial Orbit Determination of Too-short-arc Based on Kernel Density Estimation

作　　者：王歆[1,2] 李鑫冉[1,2,3] 

机构地区：[1]中国科学院紫金山天文台,南京210008 [2]中国科学院空间目标与碎片重点实验室,南京210008 [3]中国科学技术大学,合肥230026

出　　处：《天文学报》2016年第6期673-683,共11页Acta Astronomica Sinica

基　　金：国家自然科学基金项目(11373072)资助

摘　　要：采用分布估计算法,通过建立解空间的概率模型,实现了一种测角资料的极短弧初轨计算方法.在概率模型建立中,采用非参数核密度估计,无需对分布进行任何假设.不同于遗传算法、粒子群算法等进化算法,方法不仅考虑解的适值优劣,同时考虑了解的整体特性.根据基于实测数据的数值计算表明:方法在没有任何约束条件情况下对于一般观测精度仍可获得有效解.A new approach of the initial orbit determination for too-short-arc with angular measurements is implemented by building the probabilistic model in the so- lution space with the estimation of distribution algorithm. Without any assumption of distribution, the non-parametric kernel density estimation is employed in the model building. The method, unlike other evolutionary algorithms, such as genetic algorith- m and particle swarm optimization, considers the fitness as well as the characteristic of solution space. Numerical check with real observations indicates that without any constraints, the proposed technique has a good performance for the measurements of general accuracy.

关 键 词：航天器 天体力学 定轨 方法 数值 方法 统计 

分 类 号：P135[天文地球—天体力学]