基于三次卷积插值的贝叶斯滤波方法研究  

BAYESIAN FILTERING ALGORITHM BASED ON CUBIC CONVOLUTION INTERPOLATION

作　　者：王帅[1] 冯晋[2] 

机构地区：[1]长春光华学院基础教研部,长春130033 [2]哈尔滨工程大学自动化学院,哈尔滨150001

出　　处：《系统科学与数学》2015年第2期170-180,共11页Journal of Systems Science and Mathematical Sciences

基　　金：国家自然科学基金(51179039)资助课题

摘　　要：针对一类二维空间系统的状态估计模型,提出了一种用三次卷积插值方法递推估计的非线性滤波算法.仿真实例采用一个常用的非线性模型,并与粒子滤波算法进行对比分析,仿真结果表明三次卷积插值方法提高滤波估计精度,从而验证其估计一类状态估计模型解析解的可行性,其插值算法还可以推广到多维空间系统.In this paper, a nonlinear filtering algorithm applying cubic convolution interpolation to approximate state estimation model of the system. By taking the nonlinear system model as the instance and performing the comparison analysis with the particle filter, cubic convolution interpolation method can improve the filter accuracy,the feasibility of solving nonlinear model method is confirmed through simulation by using the cubic convolution interpolation experiment.

关 键 词：卷积插值 贝叶斯估计滤波 状态估计模型 

分 类 号：O212.1[理学—概率论与数理统计]