加权采样的序列蒙特卡罗滤波技术

Research of Sequential Monte Carlo Filtering Techniques Based on Weighted Sampling

在处理非线性滤波问题时,常用的几种通过模型近似或计算近似的滤波方法均在某种特定的情况下具有适用性。近年来随着计算机处理能力的快速发展,提出了序列蒙特卡罗滤波方法,因其在处理复杂的非线性和非高斯问题时表现出强大的潜力而引起广泛关注。详述了序列蒙特卡罗滤波算法的基本思想和原理,对其关键技术进行了归纳分析,并指出了该方法亟待解决的一些难点问题。

In disposing nonlinear filter question,the several filter methods,by utilizing model approximation or computation approximation,have applicability in some specific situation.In recent years,along with rapid development of computer handling ability,the sequential Monte Carlo filter method is proposed.Because the method possesses the strong potentiality in the processingcomplex non-linearity and the non-Gauss questions,it has caught people＇s attention.In this paper,the basic thoughts and principles of the sequential Monte Carlo filter algorithm are introduced.Then some critical technologies of the filter method are summarized and analyzed.Finally,some issues are pointed out which should be the focus of future study.