NLOS环境下的高斯-粒子滤波无线跟踪技术  

Wireless location technology of Gauss particle filter under NLOS environment

作　　者：毛永毅[1] 燕琳凯 

机构地区：[1]西安邮电大学电子工程学院,陕西西安710121 [2]西安邮电大学通信与信息工程学院,陕西西安710121

出　　处：《西安邮电大学学报》2016年第4期9-13,共5页Journal of Xi’an University of Posts and Telecommunications

基　　金：陕西省自然科学基金资助项目(2014JM2-6088)

摘　　要：在非视距环境下,提出一种基于高斯变换的粒子滤波跟踪算法。该算法将高斯滤波和粒子滤波相结合,利用高斯滤波产生重要密度函数,考虑最新时刻的观测数据,使采样粒子更加接近于系统状态的后验概率分布。仿真结果表明,该算法在非视距环境下具有较高的估计精度,相比于粒子滤波算法,能够减少非视距误差的影响。In nonlinear non-Gaussian environment, a novel particle filter localization algorithm based on the Gauss transform is proposed. The novel algorithm combines Gaussian filtering with particle filter, and the important density function is obtained by Gaussian filtering. The algorithm also takes into account the latest observations, and therefore the particle distribution is closer to posterior probability density function of the system state. Simulation results show that the novel algorithm has higher location accuracy in non-line of sight environment. Compared with the classical particle filter algorithm, the novel algorithm can reduce the NLOS error.

关 键 词：无线定位 非视距误差 粒子滤波 高斯变换 到达时间差 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]