运动目标状态序贯贝叶斯滤波

Sequential Bayesian Filtering for the States of a Moving Target

状态-空间模型是一个通用和宽容的模型,它包含了描述系统所要求的所有相关信息。贝叶斯滤波方法是解决状态估计问题的一般框架。当有新数据可用时,序贯滤波方法可以实现估计和更新待估参量。对于运动目标状态跟踪问题,利用状态-空间模型进行建模,并采用序贯贝叶斯方法进行处理,如扩展卡尔曼滤波、unscented卡尔曼滤波和质点滤波。该文给出了这3种方法的仿真结果,从而实现目标状态跟踪。

State-space model is a versatile and robust model.It includes all relevant information required to describe the system under investigation.Bayesian filtering provides a general framework for dynamitic estimation problems.Sequential filtering is used to estimate and update the unknown parameters of a system as data become available.State-space model and sequential Bayesian filtering are used for tracking the states of a moving target.This paper gives the result of three Bayesian filtering methods,namely extended Kalman filtering,unscented Kalman filtering and particle filtering,for tracking the states of a moving target.