DEVELOPMENT OF RFID POSITIONING SYSTEM USING NEURAL NETWORK MODEL

摘要：

　　In this paper, an overview of the RFID system which aims to measure position of RF tags covering wide area is presented.We propose a measuring method which uses relation between reaction strength from multi-antennas and RF tags 2D coordinate.We put antennas at intervals of 1.0 m to 2.0 m to read RF tags by multiple antennas and specify RF tags position by the reaction pattern.To increase the accuracy of measuring, we adapted neural network modeling and we tested various conditions to inspect the best model which minimize error of the measurement.Operation tests had been conducted at approximately 100 m2 room with 11 antennas and dynamic objects and static objects were tested.As a result we confirmed that our RFID system can measure RFID tags within 0.4 m error for dynamic objects and 0.5 m for static objects.

关键词： RFID Neural Network Work Management

作者: Takuya Sekiguchi Kazutoshi Hirata Takuya Hida Yasuhiro Kajihara Yuya Ting Takuya Mukai

作者单位: Graduate school of Tokyo Metropolitan University, Tokyo 191-0065, Japan Department of Systems Design, Tokyo Metropolitan University, Tokyo 191-0065, Japan

会议类型: 国际会议

会议名称: The 12th International Conference on Industrial Management(第十二届工业管理国际会议)

会议地点: 成都

会议语种:英文

页码: 347-352

在线出版日期: 2014-09-03（万方平台首次上网日期，不代表论文的发表时间）