摘要
针对目前基于WiFi的位置指纹室内定位算法精度不高,计算消耗时间过长,实时性差的问题,提出一种改进的室内定位算法。首先利用层次聚类的算法进行预匹配,缩小定位区域,达到降低匹配数据,提高实时性的目的,然后利用贝叶斯算法进行定位。实验结果表明,所提出的定位算法在精度和定位实时性方面都有了很大的改善。
In order to solve the problems of low precision, long computation time and poor real-time of WiFi-based position fingerprint indoor location algorithm, an improved indoor location algorithm is proposed. Firstly, the hierarchical clustering algorithm is used to pre-match, reduce the location area, reduce the matching data and improve the real-time performance. Then,Bayesian algorithm is used to locate. The experimental results show that the proposed location algorithm has a great improvement in accuracy and real-time positioning.
引文
[1] Winternitz L M B,Bamford W A,Heckler G W.A. GPS receiver for high satellite navigation. IEEE Journal of Selected Topics in Signal Processing,2009.3(4):541-556
[2]汪伦杰,廖兴宇,潘伟杰等.基于信号均值滤波+k-means+WKNN的Wifi指纹定位算法研究[J].微电子学与计算机,2017.34(3):30-34
[3]高仁强,张晓盼,熊艳,吴水平,晏磊.模糊数学的WiFi室内定位算法[J].测绘科学,2016.41(10):142-148
[4]李军,何星,蔡云泽,徐琴.基于K-means和Random Forest的WiFi室内定位方法[J].控制工程,2017.24(4):787-792
[5]曹晓祥,陈国良.一种改进的组合定权的指纹定位算法[J].测绘通报,2018.2:6-10
[6]王怡婷,郭红.基于层次聚类的WiFi室内位置指纹定位算法[J].福州大学学报(自然科学版),2017.45(1):8-15