基于网格的密度峰值聚类算法的RFID定位

Grid-based density peak clustering algorithm for RFID positioning

针对单一的一种聚类算法在处理实际定位问题时难以满足其精度需求的问题,提出一种基于网格的密度峰值聚类算法并将其应用于处理RFID定位系统中的空间数据信息,以实现对目标标签的定位。该算法结合了网格聚类算法和密度峰值聚类算法各自的特点,在保留了网格聚类算法处理大规模空间数据集的能力的同时,通过密度峰值聚类算法对网格聚类算法处理复杂聚类信息的能力进行提升,使得在处理基于RFID的室内定位问题时获得的定位效果满足实际需求。通过对3种算法的实验结果进行对比分析,可以看出算法能够提高基于RFID室内定位系统的定位精度,使得定位误差在0.128 m上下波动,具有很好的稳定性。

For a single clustering algorithm, it is difficult to meet the accuracy requirements when dealing with actual positioning problems. This paper proposes a grid-based density peak clustering algorithm and applies it to the spatial data information in the RFID positioning system to achieve the target tag positioning. The algorithm combines the features of the grid clustering algorithm and the density peak clustering algorithm. While retaining the grid clustering algorithm′s ability to handle large-scale spatial data sets, it also uses the density peak clustering algorithm for the grid clustering algorithm. The ability to handle complex clustering information is enhanced, so that the positioning effect obtained when dealing with RFID-based indoor positioning problems meets actual needs. Through comparative analysis of the experimental results of the three algorithms, the algorithm proposed in this paper can improve the positioning accuracy based on RFID indoor positioning system, making the positioning error fluctuations in 0.128 m which has a good stability.