摘要
针对室内WiFi指纹位置定位中取RSS的平均值作为其定位特征值在室内环境的复杂性和动态性不能准确地反映RSS信号真值的问题,以及卡尔曼滤波和粒子滤波算法等用于RSS信号的提取只针对线性噪声或非线性噪声中的一种,在室内动态多变、干扰复杂多样的环境下鲁棒性不理想的问题,结合卡尔曼滤波和粒子滤波,提出一种用于RSS提取的改进的粒子滤波算法。给出了算法实现的步骤,并且在不同地点不同环境条件(静态环境和动态环境)下分别进行了指纹定位在线端的数据采集实验。实验结果表明:基于改进粒子滤波的RSS提取算法的定位精度和鲁棒性均优于均值算法、卡尔曼滤波算法、粒子滤波算法等已有算法。
The main research content of this paper is the extraction of RSS in indoor WiFi fingerprint location.The mean value of RSS does not always accurately reflect the real value of RSS considering of a variety of factors for RSS measurements in the indoor environment.Meanwhile,Kalman filter or particle filter algorithm is not robustness with the ever-changing dynamic of the interior environment,considering that it is only using for linear or non-linear.This paper proposes an improved particle filter algorithm based on the Kalman filter and particle filter.The performance on WIFI indoor positioning of this value is compared with that of the mean filter,Kalman filter and particle filter under steady-state environment and dynamic environment.The experimental results demonstrate that there is a better positioning accuracy for the fusion filtering algorithm value than mean value.
出处
《测绘科学》
CSCD
北大核心
2017年第11期20-24,共5页
Science of Surveying and Mapping
基金
国家自然科学基金项目(41374011
41174010)
