摘要
室内加权K近邻(WKNN)室内定位方法的关键是指纹间距离相似度计算,现有研究忽视了信号缺失对指纹定位的影响。针对信号缺失现象,该文提出了一种接收信号强度(RSS)最小值动态补偿的室内定位方法。利用WKNN算法在3200 m^2的实验场开展定位测试,该文所提RSS最小值动态补偿的平均定位误差(ME)为2.129 m,均方根误差(RMSE)为1.78 m,相比于固定RSS补偿值,ME和RMSE分别减小了11.2%和6.5%。实验结果表明,该文所提方法可以提高室内指纹定位精度和稳定性。
The key of indoor fingerprinting localization based on weighted K nearest neighbors(WKNN)is the calculation of the distance between different fingerprints.The existing researches ignored the influence of signal lost on fingerprint localization.According to the phenomenon of signal loss,a fingerprint localization method with received signal strength(RSS)minimum dynamic compensation was proposed.Positioning test was conducted in a testbed with 3200 m^2.The proposed positioning method achieved mean error(ME)of 2.129 m,root mean square error(RMSE)of 1.78 m.Compared with the fixed RSS compensation value,ME and RMSE decreased by 11.2%and 6.5%,separately.Experimental results showed that the proposed method could improve positioning accuracy and stability.
作者
毕京学
汪云甲
纪冬华
赵文硕
BI Jingxue;WANG Yunjia;JI Donghua;ZHAO Wenshuo(Shandong Jianzhu University,Jinan 250101,China;China University of Mining and Technology,Xuzhou,Jiangsu 221116,China;Provincial Geomatics Center of Jiangsu,Nanjing 210013,China)
出处
《测绘科学》
CSCD
北大核心
2020年第12期22-27,共6页
Science of Surveying and Mapping
基金
国家重点研发计划项目(2016YFB0502102)
卫星导航系统与装备技术国家重点实验室开放基金项目(CEPNT-2018KF-03)
山东建筑大学开放实验项目(2019yzkf159)
自然资源部测绘科学与地球空间信息技术重点实验室开放基金项目(2020-3-4)
山东省高等学校青创人才引育计划项目。
关键词
室内定位
指纹定位
加权K近邻
信号缺失
接收信号强度
动态补偿
indoor positioning
fingerprint localization
weighted K nearest neighbors
signal lost
received signal strength
dynamic compensation
