最小平均积分误差粒子滤波算法的研究  

Research on Least Mean Integral Error Method for Particle Filtering

作　　者：王孟 林自豪 马俊 Wang Meng;Lin Zihao;Ma Jun(State Radio Monitoring Center Urumqi monitoring station, Urumqi, 83005)

机构地区：[1]国家无线电监测中心乌鲁木齐监测站,乌鲁木齐830054

出　　处：《数字通信世界》2018年第4期53-57,共5页Digital Communication World

摘　　要：本文基于粒子滤波算法的多普勒频差定位方法,对粒子样本贫化、定位时间相对较长进行分析,提出了应用最小平均积分误差处理的粒子滤波算法,此方法将最小平均积分误差原理应用到粒子滤波算法中,经过处理,不仅提高粒子滤波的精度,使偏离真实值的点数明显下降,也提高对不变量的估计时间。仿真实验表明:本文提出的算法在对粒子的状态估计优于传统方法,定位精度优于最小平均误差法。Based on the particle flter algorithm, the particle sampling poorly and the positioning time is long relatively, The particle filter algorithm with minimum mean integral error is introduced. This method not only improves the precision of particle flter, but also reduces the number of points that deviate from the real value, and can improve the estimation time of the invariant quantity. The method is applied to the particle flter algorithm. The proposed algorithm is superior to the traditional method in the state estimation of the particle, and the positioning accuracy is better than the minimum mean error method.

关 键 词：最小平均积分误差 进化粒子滤波 双站定位 多普勒定位 

分 类 号：TN911.72[电子电信—通信与信息系统]