摘要
室内定位已广泛应用于众多领域,目前已存在多种室内定位算法,然而多数算法的定位精度不能满足用户需求,针对基于k最近邻（k-NN）算法的室内定位精度较低的问题,提出了一种基于K-means的加权改进算法。首先通过K均方（K-means）算法处理指纹数据库,可大幅度减小测量数据的起伏波动。然后在K-NN定位算法中引入加权因子对目标节点与指纹数据库进行匹配,可有效提高匹配准确度。实验测试表明,该改进算法能有效提高定位精度40%左右,解决了基于K-NN算法的室内定位精度低的问题。
Indoor positioning had been widely used in many filed. Many indoor positioning algorithms had been proposed,but the positioning accuracy of many algorithms could not meet the requirement of users.Aiming at the problem of the low accuracy of indoor positioning based on K-NN algorithm,this paper proposed a weighted optimized algorithm based on K-means.Firstly,it used the K-means algorithm to process the fingerprint database,which could greatly reduce the fluctuations of the data.Afterwards,it used weighting factors in matching the fingerprint database,which could improve the degree of accuracy effectively.The experiment shows that the algorithm optimized proposed in this paper could improve the accuracy of positioning by 40% and resolve the problem of low accuracy of indoor positioning based on K-NN algorithm.
作者
吉彩云
袁明辉
李瑞祥
汤家森
Ji Caiyun;Yuan Minghui;Li Ruixiang;Tang Jiasen(University of Shanghai for Science and Technology,Shanghai 200093,Chin)
出处
《电子测量技术》
2018年第10期66-69,共4页
Electronic Measurement Technology
基金
上海市自然科学基金(11ZR1424900)
上海市委专项子课题(14dz1206602)项目资助
