Research And Realization Of RFID Indoor Positioning Based On ABC-BP Neural Network

Nowadays,with the rapid development of wireless technology,prompting the positioning system in the positioning speed,accuracy has been greatly improved,but also expanded its scope of application.Indoor is closely related to human life and production scene,so the study of indoor positioning algorithm has important significance.Radio frequency identification(Radio Frequency Identification,RFID)technology,is a new technology for marking and identification of object information,is one of the core technologies of the Internet of things,in the field of indoor positioning has a wide range of applications background.However,the positioning accuracy and positioning effect of RFID indoor positioning technology alone is often not ideal,in order to improve the indoor positioning effect of RFID,this paper proposes the use of neural networks and RFID combined to improve indoor positioning algorithm.Key research topics include:(1)Describe the RFID indoor positioning research background,research significance and the status quo of indoor positioning technology.The principle of RFID indoor positioning technology is introduced.The advantages and disadvantages of the algorithm are analyzed by the classification and comparative analysis of several common methods of RFID indoor positioning.(2)Describe the origin and basic principles of neural networks.This paper focuses on the algorithm flow of BP neural network and analyzes the advantages that BP neural network can bring to RFID positioning system.At the same time,BP neural network system also has obvious defects,and these disadvantages were analyzed in depth.(3)Carriy out introduction of the origin of the artificial bee colony algorithm ABC background,principle,and describes the composition of the algorithm and the algorithm flow.The operation flow and important parameter setting of artificial bee colony algorithm are analyzed.(4)Build BP neural network based on artificial bee colony algorithm optimization.Based on ABC-BP RFID indoor positioning accuracy within0.1271 m,the optimal situation is 0.0055 m,the average prediction error of0.0298 m,compared with BP system,ABC-BP positioning accuracy of the mean error is reduced by 0.0836,72.9% to enhance the accuracy.And concluded that when after artificial bee colony optimization weights and thresholds,BP neural network training effect has been strengthened,making RFID positioning effect is more excellent.