摘要
粒子滤波是基于序贯Monte Carlo仿真方法的非线性滤波算法,对基本粒子滤波算法的原理实现步骤进行了详细的介绍,进行了仿真试验。试验结果表明,粒子滤波能够很好地对非线性系统进行仿真,其估计精度要优于扩展卡尔曼滤波。由于粒子滤波算法摆脱了解决非线性滤波问题时随机量必须满足高斯分布的制约条件,并在一定程度上解决了粒子匮乏问题,近年来该算法在许多领域得到成功应用。
Sequential particle filter is a kind method of nonlinear filtering algorithm based on the Monte Carlo simulation,theories and realization steps of general particle filter algorithm are introduced in detail,and simulations are carried.The results show that particle filter is a good simulation of nonlinear systems,the estimated accuracy is superior to extended Kalman filter.The particle filter algorithm is to solve nonlinear filtering issues in which random variables must meet the conditionality of the Gaussian distribution,to some extent the degeneracy problem is solved.So the algorithm is successfully applied in many areas recently.
出处
《计算机工程与设计》
CSCD
北大核心
2009年第9期2264-2266,2269,共4页
Computer Engineering and Design
基金
国家自然科学基金项目(60675019)
第二炮兵工程学院科技创新基金项目(XY2008JJ07)
关键词
非线性滤波
贝叶斯估计
蒙特卡洛仿真
粒子滤波
目标跟踪
nonlinear filtering
Bayesian estimation
Monte Carlo simulation
particle filtering
target tracking
