基于人工神经网络的无芯片电子标签识别  

Identification of Chip-less Electronic Tag Based on Artificial Neural Networks

作　　者：车康杰 韩国瑞[1] 张文梅[1] 

机构地区：[1]山西大学物理电子工程学院,山西太原030006

出　　处：《测试技术学报》2014年第4期364-368,共5页Journal of Test and Measurement Technology

基　　金：国家自然科学基金资助项目(61271160);山西省回国留学人员项目资助项目(2012-014);教育部高校博士点基金资助项目(20121401110009);山西省自然科学基金资助项目(2012011013-3)

摘　　要：本文提出了一种基于人工神经网络的无芯片电子标签识别方法.首先利用CST对不同夹角的标签进行仿真,得到水平和垂直两个方向的散射场,然后建立识别系统的神经网络模型,并利用该模型实现对标签的识别.该方法的优点是,使用1/2的样本数据训练神经网络模型,就可以得到准确的识别结果.仿真结果表明,基于神经网络的识别方法的误差在5°的范围内.An identification method of chip-less electronic bag based on neural networks is proposed in this paper. First, the horizontal and vertical scattered fields of different tags at different angles are obtained by using CST. Second, the neural network model of recognition system is built. Then, the model is used to identify tags. The advantage of this method is that only half of sample data is needed to train the neural networks model and the accurate recognition results can be obtained. Simulation shows that the identification error of neural networks model is less than 5 degree.

关 键 词：无芯片电子标签 神经网络 函数逼近 散射场 标签识别 

分 类 号：TP29[自动化与计算机技术—检测技术与自动化装置]