Research on Big Data Processing Model of Edge-Cloud Collaboration in Cyber Physical Systems

With the improvement of the level of industrial intelligence, the traditional Cyber-Physical Systems (CPS) upload a large amount of data to the cloud computing center, which greatly reduces the quality of service (QoS) of network in production workshop, and affects the real-time interaction in the collaborative manufacturing process of each production node in the CPS. To solve this problem, a distributed big data processing model based on edge-cloud collaboration is proposed. Firstly, the distributed edge MEC server is arranged in the production workshop to disperse part of the capacity of data processing from the cloud computing center to the production node; Secondly, the deep self-encoded neural network task inference model is built, and the Neurosurge, a task unloading strategy, is used to automatically divide the network layer into parts that execute in the cloud computing center, which realize the collaborative task inference between the edge and cloud; Finally, through the experiment of tool wear monitoring, the feasibility of data processing based on distributed edge-cloud collaborative model is verified.