基于RFID路径数据的异常路径检测

发布时间：2018-05-05 23:23

本文选题：RFID + 路径数据　；参考：《计算机应用研究》2013年08期

【摘要】：RFID技术在物流、供应链管理等需要跟踪物品流动的领域的广泛使用产生了大量路径数据。路径数据描述了物品在RFID系统中的移动轨迹,路径数据中每个节点同时包含地点和时间信息,使路径数据比一般的序列数据更复杂。针对现有的序列数据异常点检测算法不适合处理路径数据的情况,对路径数据的异常路径检测进行了研究,提出适用于路径数据的扩展概率后缀树(EPST)模型和一种采用该模型检测异常路径的方法。该模型用来计算每个路径和路径数据集的相似度,在计算相似度时主要利用了路径数据的"短期记忆"性质,同时考虑了地点和时间信息对路径数据相似度的不同影响。实验表明提出的算法能够准确地检测出异常路径,并具有较低的空间复杂度。
[Abstract]:The extensive use of RFID technology in logistics, supply chain management and other areas that need to track the flow of goods has generated a large number of path data. Path data describe the moving track of items in RFID system. Each node in path data contains location and time information at the same time, which makes path data more complex than general sequential data. In view of the fact that the existing outlier detection algorithms for sequential data are not suitable for dealing with path data, the anomaly path detection of path data is studied. An extended probabilistic suffix tree (EPST) model for path data and a method for detecting abnormal paths with this model are proposed. The model is used to calculate the similarity between each path and the path data set. In calculating the similarity, the "short-term memory" nature of the path data is mainly used, and the different effects of location and time information on the similarity of path data are considered at the same time. Experiments show that the proposed algorithm can detect the abnormal path accurately and has low space complexity.
【作者单位】：江苏大学计算机科学与通信工程学院;
【基金】：国家自然科学基金资助项目(60773049) 江苏省科技型企业创新基金资助项目(BC2010172) 江苏大学高级专业人才科研启动基金资助项目(09JDG041) 高校博士点基金资助项目(20093227110005)
【分类号】：TP391.44

【参考文献】