Research And Verification Of Key Technologies Of Edge Gateway Systems For Industrial Manufacturing

With the continuous improvement of the Internet of Things technology,the industrial manufacturing field has developed rapidly.In order to solve the problem of the access management of the massive devices and the storage and process of the explosive growth of data,a new technology is needed to share the load pressure of the core network.The edge gateway system uses edge computing mode,physically close to the network edge side that near the data source,and combines the communication-computing-caching(3C)capability to provide edge intelligence services.Therefore,the research of key technologies of edge gateway systems for industrial manufacturing is valuable and necessary.The edge computing mode used as the key technology has applied in business perception,distributed processing,autonomous control,fault tolerance and privacy protection.The main works of this thesis are as follows:(1)An edge gateway system for industrial manufacturing is designed.The system integrates various heterogeneous network communication protocols such as Bluetooth,WiFi,Ethernet,OPC,Modbus,and so on.It can communicate with industrial devices of different protocols.The edge gateway system performs protocol conversion and data encapsulation according to the data attributes,and then centrally merges,classifies,and pre-processes the data received,and then connects to the Internet through a standardized API to the remote network control platform service system for data management,analysis,and sharing.The intelligent processing of data is realized.At the same time,the gateway system can centrally manage industrial equipments,and realize the intelligent control of the terminal with communication-computing-caching(3C)fusion.(2)An overhead-optimization task scheduling algorithm for edge gateway system is proposed.The algorithm considers the three factors of computing resource consumption,energy consumption and time consumption of the edge gateway,combining with the edge computing mode,aiming to minimize energy and time overhead,which is suitable for running delay-sensitive tasks in resource-constrained scenes.The algorithm uses a heterogeneous distributed network environment as the system model to calculate the overhead of locally executing tasks and offloading execution tasks,and then formulates the task scheduling strategy of the entire system.The simulations are conducted and compared with the task scheduling with randomly selected device and the cross-entropy-based optimization scheme.It is proved that the proposed task scheduling strategy can effectively reduce the system overhead and has great practicability.(3)Finally,an edge gateway system is built for research and verification,it uses Raspberry Pi as the edge node and equipped with various wireless communication module,scanning and managing t terminal devices in the precinct.The heterogeneous terminals can transmit video,voice,data and other services according to requirements.After receiving the data,the edge node preprocesses the data and sends it to the data center of the server.When the terminal sends a service request,the edge node decides whether to accept the request and allocate corresponding resource,then feeds the result back to the terminal devices.The test results show that the edge gateway system designed for industrial manufacturing can effectively support the management of terminal equipment and massive data,and achieve the expected design goals.