An RFID tag relative position positioning method based on depth learning

Abstract

The invention discloses an RFID tag relative position positioning method based on depth learning. The method comprises the following steps: S1, arranging a series of RFID tags uniformly and horizontally according to a random order, keeping an RFID antenna moving parallel to the RFID tags, and continuously reading phase information data and signal strength information data between the RFID tags byusing the RFID antenna; S2, setting a data collation structure, based on the data collation structure, making the phase information data and the signal strength information data into training data, and training the training data by using a depth neural network to initially obtain the relative positional relationship between all the RFID tags; S3, sequencing the relative positional relationships among the RFID tags by a specified sequencing algorithm, and determining the final relative positional relationships among all the RFID tags; the method of the invention can accurately analyze the arrangement order of a series of densely arranged RFID tags.

Description

technical field [0001] The invention belongs to the technical field of machine learning and the Internet of Things, and in particular relates to a mutual positioning technology of RFID (Radio Frequency Identification, radio frequency identification), especially a method for relative position positioning of RFID tags based on deep learning. Background technique [0002] Most of the current positioning research is still aimed at the absolute position. However, in many situations, the relative position of the object is more important than its absolute position. For example, in the library, librarians find misplaced books or For archives, we need to obtain the current order of books on the bookshelf rather than their absolute coordinate values; on the conveyor belt, we only need to know the relative relationship between the front and back of the object, and do not need to know the absolute coordinate value of the object. The same is true for the sorting of items in some warehous...

AI Technical Summary

Problems solved by technology

[0005] The main purpose of the present invention is to provide a method for relative position positioning of RFID tags based on deep learning. For problems with relatively low rates, the specific technical solutions are as follows:

Images