Application Research Of KubeEdge Architecture In Industrial Field

On December 10,2020,the first China 5G+ Industrial Internet Conference was held.General Secretary Xi Jinping sent a congratulatory letter.As an important part of the "new infrastructure",the Industrial Internet has become an important support for my country’s upgrade from a manufacturing power to a manufacturing power.As a new research direction that has received widespread attention,edge computing can provide faster service response and reduce network overhead by placing computing and storage on the end close to the device.This paper studies in detail the key issues in the coexistence of multi-device management and video image analysis scenarios that currently exist in the industrial Internet application of edge computing,and completes the following innovative workBased on the Kube Edge architecture,it uses a lightweight and high-performance container management engine to share the pressure of cloud computing,implement edge computing,intelligentize edge devices,and realize cloud-side collaboration and edge autonomy.At the same time,a lightweight engine is used for data and picture transmission,which can respond to abnormalities in a timely and rapid manner.Using the Tensorflow framework,a helmet detection algorithm based on videos or pictures is realized through the helmet detection model(YOLOv3)and the tracking algorithm(KCF).Multiple Raspberry Pis are deployed on the end device,and different Raspberry Pis are connected to multiple sensors,and the industrial site is monitored at the same time.The data obtained through the monitoring is centralized and transmitted to the local server,which is packaged and processed by the server and then fed back to the cloud platform.Real-time display on the front-end and Huawei IoT platform.