基于几何聚类指纹库的约束KNN室内定位模型

发布时间：2018-04-26 07:58

  本文选题：室内定位 + 聚类指纹库　； 参考：《武汉大学学报(信息科学版)》2014年11期

【摘要】：针对室内环境基于RSSI定位不稳定问题,提出了以几何信息改进基于指纹库的KNN定位算法。根据室内几何布局建立了聚类指纹库,提出了表征点位几何特性的点散发性强度(geometric strength of sporadic,GSS)概念。利用最邻近样本点的GSS判别移动终端所在参考点RP控制网结构以动态选择KNN关键参数K,构建最佳多边形为约束准则自适应选取后K-1个邻近点,建立了基于几何聚类指纹库的约束加权KNN室内定位模型。结果表明,改进后定位模型可以更好地估计终端位置信息,其中几何聚类指纹库是改善定位准确性的关键,约束KNN能够有效地提高室内定位精度。
[Abstract]:Aiming at the instability of indoor environment location based on RSSI, an improved KNN location algorithm based on fingerprint database is proposed. A cluster fingerprint library was established according to the indoor geometric layout, and the concept of point radiometric strength of sporadicus was proposed. The GSS of the nearest sample points is used to judge the RP control network structure of the reference point where the mobile terminal is located in order to dynamically select the key parameters of KNN, and to construct the best polygon as the constraint criterion to adaptively select the K-1 adjacent points. A constrained weighted KNN indoor location model based on geometric clustering fingerprint database is established. The results show that the improved location model can better estimate the terminal location information, in which the geometric clustering fingerprint database is the key to improve the location accuracy, and the constrained KNN can effectively improve the indoor positioning accuracy.
【作者单位】： 中国矿业大学国土环境与灾害监测国家测绘局重点实验室;
【基金】：国家863计划资助项目(2013AA12A201) 江苏高校优势学科建设工程资助项目 中央高校基本科研业务费专项资金资助项目(2013RC16) 新世纪优秀人才支持计划资助项目(NCET-13-1019)~~
【分类号】：TN95
，

本文编号：1805145