非线性最优滤波采样计算方法述评  

Review of sampling-data analysis methods in nonlinear optimum filtering algorithm

作　　者：郭凌云[1] 赵文丽 丁国强[1] 张志艳 

机构地区：[1]郑州轻工业学院电气信息工程学院,河南郑州450002 [2]河南省轻工业职工大学机电工程系,河南郑州450002

出　　处：《郑州轻工业学院学报（自然科学版）》2013年第5期78-84,共7页Journal of Zhengzhou University of Light Industry:Natural Science

基　　金：国家自然科学基金联合基金项目(U1204603);郑州轻工业学院博士基金项目(2011BSJJ00048)

摘　　要：基于贝叶斯估计理论基础,从减小最优滤波算法计算量、提高计算效率的角度综述了GaussHermite滤波、扩展Kalman滤波、Sigma点Kalman滤波算法中的确定性采样计算方法以及粒子滤波算法中随机样本点粒子计算方法,指出:针对粒子滤波算法中采样函数设计、重采样技术、高斯近似法与粒子滤波法的有效融合来设计研究新型高效高精度粒子最优滤波算法,将成为未来Bayesian最优滤波理论方法研究的重要领域和发展方向,而构建区间粒子最优滤波理论算法不失为粒子滤波理论算法研究的新思路.Based on the Bayesian parameters estimation theory,prospecting from the reducing amount of calculation and improving computational efficiency,the deterministic sampling methods such as the GaussHermite filtering,extended Kalman filtering,and Sigma-points Kalman filtering algorithms and the random sampling methods of particles filtering algorithms were reviewed,the research field and development direction of Bayesians optimal theory were presented,which could design the new efficient and high-precision particle filtering algorithm from the sampling function design,resampling technique and Gauss approximate method. At the same time,designing the box particle filtering algorithms became the new idea of the particle filtering algorithm.

关 键 词：贝叶斯滤波 卡尔曼滤波 确定性采样 随机性采样 粒子滤波 

分 类 号：TN957.51[电子电信—信号与信息处理]