文中在非线性贝叶斯理论的基础上介绍了粒子滤波（PF）算法及其改进算法UPF算法。利用小波变换良好的去噪能力，提出了一种结合小波变换的UPF改进算法，该算法将小波变换的去噪原理应用于粒子滤波，用于降低重要性权值的方差，从而达到提高粒子滤波的精度和稳定性的目的。仿真结果表明改进的UPF算法与原UPF算法相比，收敛速度更快，滤波精度更高，因而是一种新型的粒子滤波改进算法。

This paper introduces a particle filter algorithm and its improved algorithm-UPF on the basis of non-linear Bayesian theory. The wavelet transform has a good property of denoising. An improved UPF algorithm incorporated with wavelet transform is given in this paper, which applies the denoising principle of wavelet transform to particle filters so as to reduce the variance of importance weights and improve the accuracy and stability of particle filters. Simulation result shows that the improved UPF algorithm is not only more accurate but also has higher rate of convergence.

TN958