面向物流仓储的RFID系统数据清洗策略研究

发布时间：2018-04-13 11:56

本文选题：物流仓储 + RFID　；参考：《武汉理工大学》2015年硕士论文

【摘要】：由于射频识别(Radio Frequency Identification,RFID)技术快速、自动化识别的特点,应用于物流仓储行业将极大地提高货物信息的采集速度和物流作业效率。然而阅读器采集到的原始数据的存在大量冗余,给服务器处理数据带来巨大压力,限制了RFID技术的推广与普及。因此,如何对系统产生的海量冗余数据进行过滤清洗成为RFID技术推广需要解决的重要问题之一。本文首先介绍了RFID中间件技术的功能及意义,接着分析了面向物流仓储的RFID系统的优势,然后重点研究了RFID系统采集到的数据流的特性,包括数据流特点、数据的冗余类别及RFID有效数据的判断依据。最后详细分析了现有RFID冗余数据过滤清洗方法的特点,并分析了这些方法的优势和不足之处。在大型物流仓储的RFID管理系统中,阅读器和电子标签一般分布比较密集,冗余阅读器是影响系统性能指标的基本问题之一。在RFID系统中,判别并阻止冗余阅读器上传冗余数据可减少RFID系统能耗、系统通信量,降低系统需要处理的标签数据。文中分析了RRE、LEO、I-RRE及MRRE等几个常见的冗余阅读器清除算法的优缺点。结果表明RRE、LEO和I-RRE算法虽然能达到一定的去冗余效果,但这几个算法依赖于阅读器的读写顺序,需要对标签进行读写操作,且存在漏判问题。MRRE算法不需要阅读器对标签写入信息,只需要通过查询获取系统拓扑结构,通过矩阵运算判决阅读器是否冗余,但是该算法同样存在漏判问题。本文针对MRRE算法存在的漏判问题做出改进,提出改进的基于中间件的阅读器去冗余算法(I-MRRE),实验仿真结果表明改进算法I-MRRE的冗余阅读器检测率比MRRE提高了17.6%左右。通过冗余阅读器清除算法并不能完全杜绝冗余数据的产生,非冗余阅读器的射频交叉区域中标签数据的读取同样会产生大量冗余。为了更加快速和高效地过滤标签冗余数据,并减少系统通信量,文中利用提出的I-MRRE算法,设计单阅读器和多阅读器数据冗余过滤方法,并采用哈希表进行数据过滤。本文最后设计实验仿真软件,模拟物流仓储的RFID标签阅读器几个场景的数据采集过程,实验仿真证明设计的冗余数据过滤策略能够有效减少冗余数据的上传,降低系统通信量,减轻中间件的数据处理过程的负担。
[Abstract]:Due to the rapid and automatic identification of Radio Frequency Identification (RFID) technology, the application of RFID in logistics storage industry will greatly improve the speed of cargo information collection and the efficiency of logistics operations.However, there is a great deal of redundancy in the raw data collected by the reader, which brings great pressure to the server to process the data, which limits the popularization and popularization of RFID technology.Therefore, how to filter and clean the massive redundant data generated by the system has become one of the important problems to be solved in the popularization of RFID technology.This paper first introduces the function and significance of RFID middleware technology, then analyzes the advantages of RFID system for logistics warehousing, and then focuses on the characteristics of data stream collected by RFID system, including the characteristics of data flow.The redundant categories of data and the judgment basis of RFID effective data.Finally, the characteristics of the existing RFID redundant data filtering and cleaning methods are analyzed in detail, and the advantages and disadvantages of these methods are analyzed.In the RFID management system of large logistics warehousing, readers and electronic tags are generally distributed densely, and redundant readers are one of the basic problems affecting the performance of the system.In RFID system, discriminating and preventing redundant readers from uploading redundant data can reduce the energy consumption, system communication and label data that the system needs to deal with.In this paper, the advantages and disadvantages of several common redundant reader removal algorithms, such as RREOO I-RRE and MRRE, are analyzed.The results show that the RRENLEO and I-RRE algorithms can achieve a certain degree of redundancy, but these algorithms depend on the reading and writing order of the reader, and need to read and write the tags, and there is a problem of missing judgment .MRRE algorithm does not need the reader to write information to the tags.We only need to obtain the topology of the system by querying and decide whether the reader is redundant or not by matrix operation, but the algorithm also has the problem of missing judgment.In order to solve the problem of missing judgment in MRRE algorithm, this paper proposes an improved middleware based reader deredundancy algorithm (I-MRREN). The experimental results show that the detection rate of redundant reader based on the improved algorithm I-MRRE is about 17.6% higher than that of MRRE.Redundant reader removal algorithm can not completely eliminate the generation of redundant data, and the reading of tag data in the RF cross region of non-redundant reader will also produce a lot of redundancy.In order to filter label redundant data more quickly and efficiently, and reduce system traffic, a single reader and multiple reader data redundancy filtering method is designed by using the proposed I-MRRE algorithm, and the hash table is used to filter the data.In the end of this paper, the experiment simulation software is designed to simulate the data acquisition process of several scenes of RFID tag reader in logistics warehouse. The experimental results show that the designed redundant data filtering strategy can effectively reduce the upload of redundant data and reduce the system communication.Lighten the burden on the data processing process of the middleware.
【学位授予单位】：武汉理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP391.44

【参考文献】