基于TL-LFL的UWB室内定位算法  

UWB Indoor Positioning Algorithm Based on TL-LFL

作　　者：刘文 赵旭[1] 李连鹏[1] 刘福朝[1] 褚昕悦 代牮 LIU Wen;ZHAO Xu;LI Lian-peng;LIU Fu-chao;CHU Xin-yue;DAI Jian(Beijing Information Science&Technology University,Beijing Key Laboratory of High Dynamic Navigation Technology,Beijing 100192,China)

机构地区：[1]北京信息科技大学,高动态导航技术北京市重点实验室,北京100192

出　　处：《仪表技术与传感器》2022年第11期82-87,共6页Instrument Technique and Sensor

基　　金：国家重点研发计划课题(2020YFC1511702);北京学者计划;北京市自然科学基金(4214071)。

摘　　要：针对三边定位算法定位精度不高以及传统指纹定位算法工作量大的问题,提出了一种基于TL-LFL的室内定位算法,主要分为2个阶段:离线阶段利用限幅-滑动滤波方法对原始测距信息进行预处理,然后利用克里金插值法建立指纹数据库;在线阶段首先利用三边定位算法确定粗定位坐标和动态区域S,然后采用K-Means算法对S区域内指纹点进行聚类,找出与在线实测距离信息最近的聚类中心,最后结合粗定位坐标根据离散程度加权进行实时定位。结果表明:该算法在减少数据采集工作量的同时有效提高了定位精度,平均定位误差为0.2 m。Considering the problems that the positioning accuracy of trilateral positioning algorithm is low and the traditional fingerprint positioning algorithm has a heavy workload, a TL-LFL-based indoor positioning algorithm was proposed, which mainly included two stages.The offline stage used the limiting-sliding filtering method to pre-process the original ranging information, and then established the fingerprint database using Kriging interpolation.In the online stage, firstly, the coarse positioning coordinates and the dynamic region S were determined using the trilateral positioning algorithm, then the K-Means algorithm was used to cluster the fingerprint points in the S region to find the nearest cluster center with the online measured distance information, and finally the coarse positioning coordinates were combined with the real-time positioning weighted according to the dispersion degree.The results show that the algorithm effectively improves the positioning accuracy while reducing the data collection workload with an average positioning error of 0.2 m.

关 键 词：室内定位 UWB 三边定位算法 位置指纹定位 KRIGING K-MEANS 

分 类 号：TP301[自动化与计算机技术—计算机系统结构] TN925[自动化与计算机技术—计算机科学与技术]