In order to enhance the management level of electronic archives, an electronic archives classification management system based on single tag radio frequency identification (RFID) is designed. Use RFID to build the system framework, realize the storage and transmission of RFID data through the central database and terminal manager module, and use four modules, including file management, file management, development and utilization, and system maintenance, to complete the daily operation and maintenance of electronic file classification management; The fuzzy clustering algorithm is introduced to extract the information entropy of electronic archives data, the association rules are used to achieve data fusion and autonomous scheduling, the feature decomposition data operation state information, and the neural network classifier is established to classify electronic archives. The experimental results show that the system achieves high load balancing, low CPU utilization, and has the advantages of high accuracy and fast response rate in classification management.