Design Of Digital MES Quenching System

With the increasing demand for smart manufacturing in traditional industrial production,smart factories have become a key model for leading manufacturing innovation.As an indispensable advanced manufacturing system in smart factories,MES provides enterprises with efficient and flexible management solutions.In-depth study of MES is of great significance to promote industrial intelligent transformation.This article is based on the design of MES,relying on the actual development project of a machinery manufacturing company,and aiming at the problem of low intelligence of the quenching production line in the workshop,analyzed and designed the function and architecture of the MES of the machinery factory,and carried out quality prediction on the quenching production parameters.The main work done in this paper is as follows:(1)Combining the products and production characteristics of a certain factory,the management process,production characteristics and status of the factory are studied,and the overall demand for the MES of the machinery manufacturing factory is determined,and the overall system demand framework is built.(2)According to the business requirements of the machinery manufacturing plant,determine the functional framework and functional module design of the MES,including eight modules such as file management,material management,production management,system management,statistical analysis,equipment monitoring,production line information and enterprise information,At the same time clarify the implementation logic of each module,and analyze the four modules of file management,material management,production management,and system management in detail.Finally,the main modules in the MES are fully realized and displayed.(3)In order to adjust and optimize the production parameters in time and reduce the waste of resources in the process of production test,this paper reasonably uses the quenching production parameters,carries on the data acquisition and preprocessing,and marks each group of data with "qualified products" and "unqualified products",and divides them into training set and test set.Light BGM and XGBoost machine learning models are used for comparative analysis.Because XGBoost algorithm is accurate and fast in processing multi feature and few sample data,XGBoost algorithm is selected to design a practical quality prediction model to realize data-based production and lean manufacturing.