Design And Application Research Of H Company MES System Based On Machine Vision

Under the urgent needs of my country's implementation of the "Made in China 2025" strategic plan and the construction of the industrial Internet,the construction of digital,informatized,and intelligent factories has become the development trend and inevitable trend of various industries.This article combines the current production status of H company in the waste home appliance dismantling and processing industry in the field of domestic resource recycling and the improvement and upgrade needs of enterprises under my country's strategic planning,and carries out the design of an automatic data acquisition system based on machine vision technology,and combines the related theories of MES system.And technical requirements,completed the relatively simple MES system design and application research for H company.This paper aims to realize the real-time collection of production process data and the transparent information management between various departments of the enterprise,and transform the traditional bottom-level data information feedback to the upper-level mode to the upper-level actively absorbing the bottom-level data mode.The main research contents are as follows:First,by analyzing the industry characteristics and process flow of H company,as well as analyzing the production status of the company and the functional requirements of the company's information process and management,it is found that the information flow between the various departments of H company is not smooth and inaccurate.There is a "black box" management between the upper and lower layers.Therefore,the MES system scheme is proposed and a targeted overall design is carried out,and the key technologies required for it are analyzed in detail;Second,design the automatic collection system of various data in the production workshop of H company through the machine vision technology,which is divided into the object recognition module and the station motion analysis module.The Open CV algorithm and the pp-YOLO algorithm are used to design the pipeline object recognition module.At the same time,the KNN algorithm and SVM algorithm are used to design the classifier of the working action analysis module.Third,through the comparison and analysis of the experimental results,the pp-YOLO algorithm is finally used to design the object recognition module and the SVM algorithm is selected to design the working action analysis module.Fourth,through the experimental results,it can be known that the automatic data collection system can well realize the real-time and relatively accurate collection of production process data.In addition,the Python programming language is used and machine learning is used to realize data linkage and data analysis between raw materials,production,and finished products,and forecast production and demand,and display the data required by each department in the visualization panel.The MES system designed in this paper mainly includes data verification and analysis,job planning,product quality management and data visualization modules.