基于融合聚类的蓝牙室内定位系统算法优化

K-means聚类算法可以实现对指纹库的软划分,提高定位系统的查询效率和定位精度。由于K-means算法聚类中心选择和聚类数设定的随机性,使其稳定性较差,影响定位系统的性能,在此提出采用融合聚类的方式对K-means算法进行优化。采用基于密度峰值的聚类算法得到指纹库中每一个指纹点的局部密度和局部距离,然后计算综合决策量γ;选取跳跃点前的前k个点作为K-means算法的初始聚类中心,同时确定最佳聚类数k。试验结果表明,融合聚类算法相较于传统K-means算法定位误差在1.5 m内的概率提高了约9%,定位系统的定位精度得到明显提高。

Algorithm Optimization of Bluetooth Indoor Positioning System Based on Fusion Clustering

The K-means clustering algorithm can realize the soft division of the Radio Map and improve the query efficiency and positioning accuracy of the positioning system. Due to the randomness of clustering center selection and clustering number setting of K-means algorithm,its stability is poor and affects the performance of positioning system.The paper proposes to optimize the K-means algorithm by means of fusion clustering. The local density and local distance of each fingerprint point in the Radio Map are obtained by using the clustering by fast search and find of density peaks,and then the overall decision amount γ is calculated. The first k points before the jump point are selected as the initial clustering center of K-means algorithm,and the optimal clustering number k is determined.The experimental results show that compared with the traditional K-means algorithm,the probability of location error of fusion clustering algorithm is increased by about 9% within 1.5 m,and the positioning accuracy of the positioning system is significantly improved.