| The traditional control method for supply-oils of color offset presses is usually based on a single computer system,which has many inconveniences in realizing remote control and centralized management.Based on the new system architecture of cloud-edge-end,the development of a new type of ink volume control technology for multi-color offset presses is of great significance for further improving the technical level of domestic printing equipment.This thesis designs an cloud control system for supply-oils of multi-color offset press,and studies its key technologies.This system is mainly divided into three parts: ink volume control terminal,image acquisition equipment and edge computing platform.Based on the cloud,edge,and terminal system architecture,the terminal layer implements ink volume control and print image acquisition tasks,the edge layer implements functions such as image recognition and model training,and the cloud layer manages and schedules all nodes in the edge layer online.In order to reduce equipment cost,the ink volume control terminal adopts the master-slave architecture mode with STM32 as the control core.First of all,in order to meet the stability,distance and speed requirements of data transmission,the master and slave are connected through 485 bus and adopt Modbus protocol.Secondly,the remote online monitoring and visual remote control of ink volume based on smart phone are developed,the mobile APP control method based on Gizwits platform and MQTT protocol and the Lab VIEW control method based on TCP/IP protocol are used,and the LWIP protocol stack transplanted in the main station is used.The provided RAW API programming interface implements web page embedding technology.Finally,the ?C/OS-? operating system is transplanted in the master station to ensure the orderly progress of multiple tasks of the system.At the same time,the i.MX6 ULL processor is used in the technical scheme of the printed image acquisition equipment.The Wi-Fi module of the device is connected to the local area network to communicate with the edge nodes wirelessly.The fast charging module is selected to realize the functions of fast charging and low battery reminder.Time mem low power consumption mode guarantees battery life,adopts V4L2 framework to drive development and application programming of the camera,designs compound light source to realize the collection of anti-counterfeiting images,and realizes the functions of auto focus,optical magnification and automatic compensation of light intensity.This system scheme uses Raspberry Pi 4B as the core of the edge computing platform,uses the Py Qt framework for visual interface design,Open CV and pyzbar libraries directly collect and identify printed anti-counterfeiting images,and Open CV and YOLOv5 algorithm for defect detection.After experimental tests,the cloud control system for supply-oils designed in this thesis has the characteristics of high precision,low delay and feedback adjustment;the printed product image acquisition equipment can meet the real-time online monitoring of printed product quality and the functions of anti-counterfeiting image acquisition and reading;The edge computing platform has better results in the identification of printed anti-counterfeiting images and the detection of printed product defects. |
PDF Full Text Request