Design Of Product Quality Information Management System Based On Production Data Analysis

In recent years,with the rapid development of China’s economy and technology,steel pipe enterprises are also moving in the direction of automation,digitalization and intelligent production.However,at present,steel pipe enterprises still need to manually determine the rating of steel pipes through a number of process posts,and enterprises have not been able to achieve the quality of steel pipe information management system.To solve these two problems,this paper designed and implemented a set of product quality information management system based on production data analysis,and realized the quality information management system for steel pipe production process integrating steel pipe rating,comprehensive display,data entry,data traceability and defect management.The contents and innovations of this paper are as follows:(1)Combined with the production status of a domestic steel pipe enterprise,the demand of the enterprise was analyzed,and the problem that the enterprise manually rated the steel pipe required multiple process positions to judge and the overall demand for the lack of steel pipe quality information management system was clarified.The design scheme of abnormal detection and rating of steel pipe quality data and steel pipe quality information management system is obtained.(2)In the steel pipe rating module of the system,this paper proposes a Deep Autoencoding Gaussian Mixture Model with Attention Mechanism(DAGMM-AM)to achieve abnormal detection and rating of steel pipe quality data.The addition of attention computation can improve the efficiency of the model to collect important features from the data,while suppressing redundant information in the original data.The model combines the whole training process,from small iterative loops to the whole network,and uses a mixed loss function to conduct end-to-end joint training,which helps the autoencoder get rid of local optimality in the training process.The DAGMM-AM algorithm is feasible to grade steel pipes after testing.(3)After analyzing the production status and demand of steel pipe enterprises,focusing on the two core objectives of steel pipe quality data analysis and quality information management,the whole system is divided into multiple functional modules of steel pipe rating,comprehensive display,data entry,data traceability,defect management,quality report and authority management.In order to solve the problem of manual steel pipe rating,DAGMM-AM algorithm is used to automatically grade steel pipe rating module in the system,and other functional modules are designed to solve the problem of lack of steel pipe quality information management system.Through the combination of steel pipe rating and quality information management,the steel pipe quality information management system based on the analysis of steel pipe production quality data is realized,and the system functions are explained according to the renderings on the page.