基于RFID位置语义的室内移动轨迹聚类算法

面向室内空间的移动轨迹聚类有利于发现室内热点和用户移动模式.针对室内环境在定位技术、距离度量等方面的特殊性,充分考虑室内移动轨迹的空间和语义特征,提出一种基于无线射频识别(radio frequency identi-fication,RFID)位置语义的室内移动轨迹聚类方法.该方法对原始轨迹提取特征点,可简化轨迹以降低算法时间复杂度;从空间形状和位置语义2个方面加权计算轨迹相似度,其中,空间相似度通过定义适用于室内三维空间的距离函数来计算,语义相似度计算基于最长公共子序列思想,并引入移动对象在轨迹点的到达时间和停留时间;利用线性表存储轨迹相似度,采用改进的层次聚类方法对移动轨迹进行聚类.实验结果表明,该方法能够有效地进行室内轨迹聚类并具有较高的效率.

Clustering algorithm for indoor moving trajectory based on RFID location semantics

Indoor moving trajectory clustering is beneficial to find indoor hotpots and user' s mobility patterns. For the parti-cular aspects of positioning technology and distance measurement, this paper proposes a clustering algorithm for indoor mov-ing trajectory based on RFID location semantics, which takes into full account the spatial and semantic features of indoor moving trajectory. This method extracts critical points of trajectories to reduce the time complexity of algorithm. Then it sets a weight parameter to measure trajectory similarity from two aspects of spatial shape and location semantics. While spatial similarity calculation is based on the definition of distance function which applies to indoor three-dimensional space. Seman-tic similarity calculation is based on longest common subsequence and using the moving object' s arrival time and stay dura-tion at a track point. At last trajectory similarity is stored with linear table, and an improved hierarchical clustering method is adopted to find trajectory clusters. Experiments indicate that this method works effectively and improves the efficiency of indoor trajectory clustering.