室内环境下基于IMM-EKF算法的移动目标定位

发布时间：2018-01-08 15:15

本文关键词：室内环境下基于IMM-EKF算法的移动目标定位　出处：《计算机研究与发展》2014年11期 　论文类型：期刊论文

【摘要】：如何在视距(line-of-sight,LOS)与非视距(non-line-of-sight,NLOS)混合的室内环境中提高移动目标定位的精度,这是一项富有挑战性的工作.移动目标在室内环境移动时,其与传感器网络节点之间的信号传播在LOS与NLOS之间随机切换,导致移动节点定位精度下降.提出一种交互式多模型-扩展卡尔曼滤波(interactive multiple model-extended Kalman filter,IMM-EKF)定位算法.根据LOS?NLOS环境下不同的测距误差特性,在IMM框架中采用2个平行的卡尔曼滤波器(Kalman filter,KF)模型对测量距离同时进行滤波,根据滤波结果和测量值计算2个模型的似然概率,模型间的转换通过Markov链实现,2个KF滤波结果加权融合后获得IMM距离估计值.在EKF定位阶段,通过位置预测和更新估计出移动目标位置.仿真结果表明,IMM-EKF算法能够有效抑制NLOS对目标定位的影响,其定位精度优于单模型算法.
[Abstract]:How in the line of sight (line-of-sight, LOS) and non line of sight (non-line-of-sight, NLOS) to improve the mobile target positioning accuracy of hybrid indoor environment, this is a challenging task. The moving target in the indoor environment of mobile, communication signal between the sensor network node between LOS and NLOS random switching, cause drop the mobile node positioning accuracy. This paper presents an interactive multiple model extended Calman filter (interactive multiple model-extended Kalman filter, IMM-EKF) positioning algorithm. According to the different characteristics of the ranging error of LOS? NLOS environment, using 2 parallel Calman filter in the IMM framework (Kalman filter, KF) at the same time filtering model on measuring distance according to the filtering result, the likelihood and calculation of the measured values of the 2 models, the transformation between models is realized through the Markov chain, 2 KF filtering results obtained after weighted fusion IMM Distance estimation. In the EKF positioning stage, the location of mobile targets is estimated by location prediction and update. Simulation results show that IMM-EKF algorithm can effectively suppress the influence of NLOS on target location, and its location accuracy is better than that of single model algorithm.

【作者单位】：东北大学信息科学与工程学院;
【基金】：国家自然科学基金项目(61273078) 中国博士后科学基金面上项目(2012M511164) 中央高校基本科研业务费专项资金项目(N130404023,N110804004) 辽宁省博士启动基金项目(20121004)
【分类号】：TP212.9;TN929.5
【正文快照】： 传统的GPS定位技术在室外环境下能够取得较好的定位效果.但在室内环境中,由于信号受到建筑物的遮挡,GPS无法实现较为精确的定位.无线传感器网络(wireless sensor network,WSN)因其自组织、易部署、低能耗和可大规模部署等特点,在室内移动目标的定位领域具有良好的应用前景.基

【参考文献】