Liquid Material Identification based on RFID Passive Sensing and Machine Learning

In order to further improve the identification accuracy of unknown liquid materials, this paper proposes a liquid material identification method based on RFID passive sensing and machine learning. Firstly, based on the principle that phase changes and received signal strength (RSS) changes are different when RFID signal passes through different liquids, the phase and attenuation constants of a specific liquid are derived and defined as its two characteristic values. Then, the weighted K-nearest neighbor algorithm (WKNN) is used to optimize the weights of training samples at different distances, while machine learning is performed on labeled liquid feature data. Finally, an experimental system based on Impinj R420 RFID reader is established and numerous experimental tests are carried out. The test results show that the proposed approach can effectively identify ten common liquids such as beer, coke, spirit and milk, with an average identification accuracy of 99.4%, which is 4.1% higher than the existing methods, implying its good theoretical significance and application value.