Data Mining Method And Application Research Of Pressure Vessel Welding Quality Control

At present,welding in the pressure vessel manufacturing process more and more important position,in order to ensure the quality of the pressure vessel,the implementation of welding quality control and management is particularly critical.With the rapid development of computer technology and database technology,the modern manufacturing industry has produced huge amounts of data in the production process.Obviously,the traditional statistical methods are in a dilemma and can not meet the development needs of the present age.In this paper,we can not extract valuable "knowledge" from the massive data in modern manufacturing industry,and combine data mining technology with the practical project of a company to control the welding quality in the manufacturing industry.In this paper,based on previous studies,focused on the pressure vessel welding quality control work,the study summarized as follows:First of all,for the pressure vessel welding quality failure phenomenon,from the quality management 5M1E(operator,machinery and equipment,raw materials,process,environment and measurement)and other aspects,the whole process,multi-faceted,multi-angle analysis of pressure Container welding quality factors,reduce the scope of welding quality anomalies.Secondly,this paper proposes a classification method based on feature selection and decision tree C5.0.Firstly,feature selection algorithm is used to reduce dimension of a large number of features,and the reduced dimension welding data is constructed by decision tree C5.0 algorithm.Welding classification prediction model,so as to find out the factors affecting the quality of welding failure.These important factors as"quality control point" and strict control.Finally,the decision tree algorithm is compared with the neural network algorithm and the Logistic regression algorithm.Experiments show that the decision tree method is superior to the neural network algorithm and the Logistic regression algorithm for the problem studied in this paper,and the decision tree analysis The results are applied to the actual welding work of the company,the welding problem has been obviously improved,and its idea and guiding ideology have certain popularization and application value in the pressure vessel manufacturing industry.