论文部分内容阅读
粒子滤波已成为研究非线性非高斯估计问题的重要方法,近年来备受关注。在递推贝叶斯估计理论框架下,分析了粒子滤波理论的核心思想、基本原理及其性能特点;接着,探讨了粒子滤波理论中建议密度的选取和重采样两个关键问题,较全面地总结了相应的选取方法和改进策略;并对粒子滤波理论在导航定位方面的应用进行介绍。对粒子滤波算法进行展望,指出一些未来的研究方向。
Particle filter has become an important method to study nonlinear non-Gaussian estimation problems and has drawn much attention in recent years. Under the framework of recursive Bayesian estimation theory, the core idea, basic principle and performance characteristics of particle filter theory are analyzed. Secondly, two key issues of the proposed density selection and resampling in particle filter theory are discussed. Summarizes the corresponding selection method and improvement strategy; and introduces the application of particle filter theory in navigation and positioning. The particle filter algorithm is prospected and some future research directions are pointed out.