摘要
针对一般粒子滤波中的退化问题,提出了一种改进的Unscented卡尔曼粒子滤波(UPF)算法。提出了最小偏度采样策略,将该策略应用于UKF算法中,以UKF方法生成建议分布并从中采样,解决了一般粒子滤波算法中以转换先验密度函数作为替代分布所引发的粒子退化等问题。将多项式重采样和分层重采样两种方法结合起来进行重采样,有效地减弱了粒子退化问题。最后,给出了非线性序列的仿真算例。理论分析和仿真结果表明,改进的UPF算法提高了滤波的稳定性和精度,算法的运行效率提高了30%。
To solve the degeneracy problem of general particle filter, an improved Unscented Kalman Particle Filter（UPF） algorithm is proposed. A new sampling strategy calledminimal skew sampling is proposed. The proposed sampling strategy is used in the UKF algorithm and a proposal distribution is generated and the samples are extracted by UKF. By doing this, some problems such as particle degeneracy caused by transition prior density function as proposal distribution are solved. A new resampiing method combining multinomial resampling with stratified resampling is proposed to efficiently solves the degeneracy problem of particle filtering. Finally, a simulation example of a nonlinear time sequence is given. The results of theory analysis and simulation show that the improved UPF algorithm improves the stability and accuracy of filtering, and the operating efficiency increases by 30%.
出处
《光学精密工程》
EI
CAS
CSCD
北大核心
2008年第4期746-751,共6页
Optics and Precision Engineering
基金
部委预研基金资助项目(No.9140A17080407DZ101)
部委"十一五"预研资助项目(No.51316060205)
