摘要
针对WiFi指纹室内定位技术中的接收信号强度的偏态分布问题,该文提出一种缩放因子自动优化的改进箱线图滤波方法。该方法利用中位数构建非对称的滤波区间,同时利用位置比对进行缩放因子的自动优化,进行了原始观测数据的粗差剔除,解决了非高斯分布噪声的影响,实现了复杂室内场景中的高精度WiFi室内定位。实验结果表明,所提出的算法可以有效地剔除WiFi指纹室内定位技术中的观测噪声,滤波算法的缩放因子参数实现了自动优化,具有更好的鲁棒性,提高了WiFi指纹室内定位的精度和可靠度。
Aiming at the skew distribution problem of received signal strength in WiFi fingerprinting indoor positioning technology,an improved box-plot filtering method with automatic optimization scaling factor was proposed.This method used median to construct an asymmetric filter interval and its scaling factor was automatically optimized through position comparison.The new algorithm removed the gross errors of the original observation data and solved the influence of non-Gaussian distributed noise,finally realized the high-precision WiFi indoor positioning in complex scenes.The experimental results showed that the proposed algorithm could effectively eliminate the observation noise in the WiFi fingerprinting indoor positioning technology,and the scaling factor parameter of the filtering algorithm was automatically optimized,which finally improved the accuracy and reliability of WiFi fingerprinting indoor positioning.
作者
张伟
连国云
花向红
王刘准
冯健
ZHANG Wei;LIAN Guoyun;HUA Xianghong;WANG Liuzhun;FENG Jian(School of Artificial Intelligence,Shenzhen Polytechnic,Shenzhen,Guangdong 518000.China;School of Geodesy and Geomatics,Wuhan University,Wuhan 430079,China;Shenzhen Ronghui Intelligent Technology Co.,Ltd.,Shenzhen,Guangdong 518000,China;Hangzhou Doudou Intelligent Technology Co.,Ltd.,Hangzhou 310051,China)
出处
《测绘科学》
CSCD
北大核心
2021年第12期61-66,共6页
Science of Surveying and Mapping
基金
国家自然科学基金项目(41674005,41871373)
广东省自然科学基金项目(2019A1515011267)
广东省教育厅重点领域专项(2021ZDZX1090)
深圳市基础研究计划项目(JCYJ20190809113617119)。
关键词
WiFi指纹室内定位
接收信号强度
偏态分布
缩放因子
箱线图滤波
WiFi fingerprinting indoor positioning
received signal strength
skew distribution
scaling factor
boxplot filtering
作者简介
张伟(1989-),男(土家族),湖南张家界人,讲师,博士,主要研究方向为多传感器信息融合无缝导航定位技术。E-mail:2012202140029@whu.edu.cn;通信作者:花向红,教授,E-mail:xhhua@sgg.whu.edu.cn。
