The Technology Research Of Massive Data Fusion For Workshop Based On Internet Of Things

This subject originates from the National Defense Basic Scientific and Research Project---A Study on creating “Local Area” Internet of Things and Data Management Platform at the Military Products field. The project brings the Internet of Things technology into the field of manufacturing data acquisition, transmission, processing and application for the military industrial enterprises so as to realize the comprehensive perception, reliable transmission and intelligent processing of information. Specific to the isomerism and other features of data acquisition in the military workshop manufacturing fields, the paper analyzes two key technologies of Internet of Things, namely, Radio Frequency Identification Technology(RFID) and Wireless Sensor Network(WSN). Based on this, the paper also profoundly studies the data cleaning and data fusion methods, favorably improves network congestion, low communication efficiency, low precision and reliability and other problems that brought by the mass data of the Internet of Things, and effectively supports the upper applications’ requirements for complete, precise and real-time data. The paper’s main contents are as follows:(1) Provides the general framework for military industrial Internet of Things manufacturing workshop’s mass data fusion system; introduces the work procedures and main modules of the system, and elaborates key technologies involved in the system.(2) Puts forward an adaptive RFID data cleaning algorithm based on sliding-window. The algorithm employs sliding window technology and binomial distribution model to calculate the appropriate window size, and adjusts the window size dynamically through the window subintervals’ monitoring results and tag dynamics, in order to correspond to the fluctuation of readers’ tag performance caused by environmental changes, thus testing the transition points in a relatively accurate manner and reducing the mistakes in data reading.(3) Puts forward a data stream and network coding-based data aggregation algorithm in Wireless Sensor Networks. The algorithm research mainly focuses on data transmission and energy consumption. The algorithm reduces data transmission through the nodes’ computing and storage capacity and data flow technology during the source nodes’ data perception. Within the cluster, according to the data packages’ relevant grades, the algorithm uses network coding and data aggregate technologies to reduce the cluster head’s extra energy consumption and increase the Internet of Things manufacturing workshop’s data package delivery rate and energy efficiency.(4) Completes the design of military production Internet of Things workshop’s mass data fusion system and realized main functions. Under the background of military products field manufacturing “local area” Internet of Things and data management platform, the paper applies the RFID data cleaning based on sliding window and data fusion algorithm based on data flow and network coding in the project, finishes the system’s demand analysis and design, and provides the system’s functional structure and realization procedures of its main functions.