基于粒子滤波和滑动平均扩展Kalman滤波的多径估计算法

发布时间：2018-03-21 01:15

  本文选题：参数估计　切入点：粒子滤波　出处：《电子与信息学报》2017年03期 　论文类型：期刊论文

【摘要】：多径干扰是高精度定位的主要误差源,估计多径参数对消除多径误差,提高导航系统定位精度具有重要意义。针对扩展Kalman滤波(EKF)在进行多径参数估计时,存在对初值敏感,估计结果在真值附近具有较大波动的缺点,该文提出一种基于粒子滤波(PF)和滑动平均EKF的多径估计算法。该算法首先利用PF得到多径参数的粗略估计值,并将该值作为EKF的初始估计值,以克服EKF对初值敏感的问题。接着对EKF的估计结果进行滑动平均,并将平均后的滤波结果作为多径参数的估计结果。仿真结果表明,改进后的多径估计算法可有效降低估计结果的波动幅度,同时解决了EKF对初值敏感的问题。
[Abstract]:Multipath interference is the main error source of high precision positioning. It is important to estimate multipath parameters to eliminate multipath errors and to improve positioning accuracy of navigation system. The extended Kalman filter is sensitive to initial values when it is used to estimate multipath parameters. This paper presents a multipath estimation algorithm based on particle filter (PF) and moving average EKF (EKF). Firstly, a rough estimation of multipath parameters is obtained by using PF. In order to overcome the problem that EKF is sensitive to the initial value, this value is taken as the initial estimation value of EKF. Then, the moving average of the EKF estimation result is carried out, and the average filtering result is taken as the estimation result of the multipath parameter. The simulation results show that, The improved multipath estimation algorithm can effectively reduce the fluctuation of the estimation results and solve the problem that EKF is sensitive to the initial value.
【作者单位】： 太原理工大学信息工程学院自动化系;北京理工大学复杂系统智能控制与决策国家重点实验室;
【基金】：国家自然科学基金(61503271,61603267) 山西省自然科学基金(20140210022-7) 复杂系统智能控制与决策国家重点实验室开放基金(900101-03910353)~~
【分类号】：TN713
，

本文编号：1641618