摘要
介绍基于粒子滤波器的非线性估计方法。采用正则化粒子滤波器来缓解粒子滤波器重采样造成的问题,改进了粒子滤波器的性能。在一种典型的非静态增长模型下比较EKF,UKF,PF和RPF的滤波性能差异。仿真结果表明,PF在滤波精度方面优于EKF和UKF,而RPF在精度和计算复杂度等方面均优于PF。
Nonlinear estimation methods based on Particle Filter (PF) are proposed. Regularized Particle Filter (RPF) is emphasized to relieve the problems caused by resampling of PF, and improve the performance of PF. The comparison of filtering performance among EKF, UKF,PF and RPF is made in a typical nonstatic model. The simulation results show that PF is better than EKF and UKF in the performance of accuracy,and the performance of RPF is better than PF in both filtering accuracy and calculating complexity.
出处
《现代电子技术》
2009年第4期141-144,共4页
Modern Electronics Technique