基于分数低阶统计模型的粒子滤波器  

A Fractional Filter Based on the Fractional Lower Order Statistical Model

作　　者：杨保海[1] 査代奉[1] 

机构地区：[1]九江学院电子工程学院,江西九江332005

出　　处：《组合机床与自动化加工技术》2010年第2期35-36,40,共3页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：国家自然科学基金(60772037);江西省教育厅科技项目(GJJ09344)

摘　　要：在alpha稳定分布噪声下,传统的基于二阶模型的粒子滤波器性能会出现退化,文章提出了一种基于分数低阶模型的分数低阶粒子滤波器(FLOS_PF),并给出了具体的运算公式。用这种方法对高斯噪声和alpha稳定分布噪声下的正弦信号分别进行仿真,仿真结果表明,分数低阶粒子滤波器在处理高斯噪声时,其结果与标准的粒子滤波器(GSPF)相当,但是在处理alpha稳定分布噪声时,分数低阶粒子滤波器的仿真结果明显优于标准的粒子滤波器。The performance of traditional particle filter would be degenerate in alpha stable distribution noise. A new fractional lower order particle filter based on fractional lower order model was presented in this paper, and the detailed operation algorithm was proposed. The sine signal embeded in gaussion noise and the alpha stable distribition noise was simulated with the new method respectively,Simulation results showed that the performance of the new method based on fractional lower order model similar to standard panicle filter in gausssion noise while outperforms the standard panicle filter in alpha stable distribution noise.

关 键 词：分数低阶 Α稳定分布 粒子滤波器 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]