摘要
粒子滤波(PF)是解决非线性、非高斯估计问题的重要方法,是用粒子及其权重来表示后验概率密度。随机粒子建议密度的选取至关重要,决定着滤波性能。对扩展H∞粒子滤波算法(PF-EHF)进行研究,并应用于惯导系统非线性初始对准的状态估计。PF-EHF算法是以扩展H∞滤波(EHF)作为建议密度的新型粒子滤波算法,它便于利用最新的观测数据更新粒子,得到优化的状态估计和方差,进而改善滤波性能。仿真结果表明,PF-EHF的定位精度高于传统的EKF。
An improved particle filter, named extended H~ particle filter （PF- EHF） was discussed in the pa- per. Its proposal distribution was generated on the basis of the extended H~ filter （El-IF）. The EHF can update par- ticles utilizing the new observations, which makes the EHF a better candidate for more accurate proposal distribution generation within the particle filter framework. And then, PF - EHF was used to estimate the INS nonlinear initial a- lignment on stationary base. The simulation results show that PF - EHF can yield better performance than EKF does.
出处
《计算机仿真》
CSCD
北大核心
2013年第4期37-40,共4页
Computer Simulation
基金
中国博士后科学基金项目(20080431094)
江苏省博士后科研资助计划项目(0801020B)
关键词
惯性导航系统
初始对准
建议密度
滤波
粒子滤波
Inertial navigation system
Initial alignment
Proposal distribution
filter
Particle filter
