论文部分内容阅读
非线性非高斯状态空间模型的最优估计问题在信号处理、自动控制、金融、无线通讯等领域具有重要的应用,粒子滤波技术通过非参数化的蒙特卡罗模拟方法来实现递推贝叶斯滤波,适用于任何能用状态空间模型表示的非线性系统,滤波精度可以逼近最优估计,其有效性已经得到各领域研究人员的极大认可,基本粒子滤波算法存在的最大问题是粒子退化,针对这一问题,对权值退化、重要性函数选取、重采样等影响粒子滤波器性能的关键技术进行深入研究。
The optimal estimation problem of nonlinear non-Gaussian state space model has important applications in the fields of signal processing, automatic control, finance, wireless communication and so on. The particle filter technology realizes the recursive Bayesian method through nonparametric Monte Carlo simulation It can be applied to any nonlinear system that can be represented by the state space model. The filtering accuracy can be approximated to the optimal estimation. Its validity has been greatly recognized by researchers in various fields. The biggest problem of the basic particle filter algorithm is particle degeneration, Aiming at this problem, the key technologies that affect the performance of particle filter, such as weight degradation, importance function selection and resampling, are deeply studied.