基于证据理论的群指纹融合室内定位方法
发布时间:2019-07-15 16:21
【摘要】:室内定位的主要挑战是室内的多径传播及非平稳信道环境,传统基于信号强度指纹的单指纹室内定位方法由于受环境变化影响较大,稳健性较差且精度较低。针对此问题,提出一种基于D-S证据理论的群指纹融合高精度室内定位方法。在建库阶段,利用室内阵列信号接收模型,首先通过计算阵列接收信号的不同统计特性构建包括信号强度、协方差矩阵、信号子空间及四阶累积量组成的群指纹库,再对群指纹进行神经网络训练获取针对每种指纹的神经网络分类器;在实测阶段,把实测数据的上述4种变换输入到训练好的神经网络分类器中,最后利用D-S证据理论对神经网络分类器的分类结果进行融合,给出最终的定位结果。仿真结果证明了算法的有效性及可行性。该算法可充分发挥指纹信息的集群效应,对噪声、多径传播等具有较好的稳健性,是一种高精度的室内定位新方法。
[Abstract]:The main challenge of indoor positioning is indoor multi-path propagation and non-stationary channel environment. The traditional single fingerprint indoor location method based on signal strength fingerprint is greatly affected by environmental changes, poor robustness and low accuracy. In order to solve this problem, a group fingerprint fusion high precision indoor location method based on D 鈮,
本文编号:2514761
[Abstract]:The main challenge of indoor positioning is indoor multi-path propagation and non-stationary channel environment. The traditional single fingerprint indoor location method based on signal strength fingerprint is greatly affected by environmental changes, poor robustness and low accuracy. In order to solve this problem, a group fingerprint fusion high precision indoor location method based on D 鈮,
本文编号:2514761
本文链接:https://www.wllwen.com/kejilunwen/xinxigongchenglunwen/2514761.html