摘要
针对室内定位指纹算法的定位精度以及实时性问题,提出一种基于核K-means和相关向量机的定位算法。该算法首先使用核K-means算法将接收信号强度进行聚类,存入指纹特征数据库,通过RVM回归对指纹数据库进行训练,算出最优拟合位置的数学模型。实验结果表明,该算法于定位实时性以及定位精度优于SVM相关定位算法。
With respect to the problem of location accuracy and real-time performance of fingerprint indoor positioning algorithm. Proposes a location algorithm based on kernel K-means and relevance vector machine. The algorithm first uses the kernel K-means algorithm to cluster Received Signal Strength data and stores it as a fingerprint database. Then the RVM regression algorithm is used to train the fingerprint data in the database. Through the training we can calculate the best fitting position of the mathematical model. Experimental results show that the algorithm is superior to the SVM-related positioning algorithm in real-time performance, superior to the SVM-related positioning algorithm in positioning accuracy.
作者
陈骁
宋安军
CHEN Xiao;SONG An-jun(College of Information Engineering,Shanghai Maritime University,Shanghai 201306)
