Research On Fingerprint Indoor Location Algorithm Of RFID Based On Improved BP Neural Network

With the development of communication technology,location-based service is more and more widely used in life,which makes indoor positioning technology more and more important.Due to the complex indoor environment and numerous disturbances,how to improve the indoor positioning accuracy is the main problem in the field of indoor positioning.Due to the rise of the Internet of things,RFID,as one of the key technologies,has been widely used in the indoor positioning field by virtue of its features of non-line-of-sight,short delay,low cost and large transmission range.Therefore,the research of indoor positioning technology based on RFID has become a hot spot in the field of indoor positioning.Among the indoor positioning methods based on RFID,the fingerprint indoor positioning algorithm of RFID based on non-ranging has strong flexibility,low hardware cost and high accuracy,which is the mainstream indoor positioning method based on RFID indoor positioning.In research on fingerprint indoor location algorithm based on RFID,it is a hot topic to improve fingerprint indoor location algorithm based on RFID with BP neural network algorithm.In order to further improve the indoor positioning accuracy,this paper is based on the following two aspects of the fingerprint indoor positioning algorithm of RFID based on BP neural network:(1)In the fingerprint indoor positioning algorithm based on RFID,readers are required to collect the RSSI value of the tag in both offline and online positioning stages.Due to the complexity of the indoor positioning environment,more RSSI values need to be collected to preprocessing to obtain an accurate RSSI value.This paper presents a hybrid filtering algorithm based on Dixon test filtering,particle filtering and kalman filtering.It can better deal with random interference and mutation interference RSSI value.The simulation results show that the hybrid filtering algorithm improves the accuracy and anti-interference of RSSI estimation.(2)In the fingerprint indoor positioning algorithm of RFID based on BP neural network,It is the key to the accuracy of indoor positioning algorithm,which is the determination of weight and threshold of BP neural network.In this paper,fingerprint indoor location algorithm of RFID of BP neural network base on fireworks explosion algorithm is proposed.It optimizes the BP neural network by using the fireworks explosion algorithm with the advantages of good overall optimization.The optimal weight and threshold were determinedand the BP neural network model was established.The model is used to complete the positioning of undetermined tags in fingerprint indoor positioning algorithm based on RFID.Finally,the simulation results show the algorithm presented in this paper has less errors to a certain extent and can improve the indoor location accuracy.