中国空间科学技术

• •    

基于最优节点样条逼近的观测数据平滑方法

郭军海   

  1. 北京跟踪与通信技术研究所!北京100094
  • 发布日期:2000-06-25

A Measurements Smoothing Method Based on Optimal Knots Spline Approximation

Guo Junhai (Beijing Institute of Tracking and Telecommunication Technology,Beijing 100094)   

  • Published:2000-06-25

摘要： 提供了一种基于最优节点样条逼近的观测数据平滑方法。它利用节点自由分布 B样条描述目标运动 ,在最小二乘准则下 ,采用非线性最优化方法 ,搜寻样条的最优节点分布 ,使逼近过程能够实时自适应目标的运动规律 ,提高逼近精度。这种方法即使在目标运动变化剧烈的情况下仍有较好的平滑和微分平滑效果

关键词: 样条函数, 平滑, 数值分析

Abstract: A smoothing method for noisy measurements based on optimal knots spline is presented.Describes the objective movement using the B spline with free knots,and in terms of the least square principle,the optimal knots distribution is searched employing the nonlinear optimization.Consequently,the approximative process is adaptively fit for the objective movement,and the approximative precision is improved.The ideal smoothing and differentiation smoothing effect can be obtained even if the object moves rapidly.